# Helena Bioinformatics - full public content corpus > Consolidated machine-readable text for every canonical URL declared in https://www.helena.bio/sitemap.xml. > Latest recorded content update: 2026-10-16. > Canonical HTML pages and their cited sources remain authoritative. Pages included: 18 --- ## Page: Helena Bioinformatics | Clinical Genomics Infrastructure Source: https://www.helena.bio Canonical: https://www.helena.bio Description: Helena Bioinformatics builds evidence-traceable software and infrastructure for clinical genomics and biomedical research. # Intelligence for Genomics We build the products genomic science needs to scale. ## From a VCF to a clinical report Folklore brings classification, evidence, phenotype and prioritisation into one case view. A qualified professional reviews the final interpretation. - Panels, exomes and whole genomes - Classification, evidence and prioritisation - Model-assisted interpretation under clinical review Paused until Folklore is in focus Folklore Intro, video 1 of 4. Swipe left or right to change the videos. Use the sound button to turn audio on or off. Paused until Undertone is in focus Undertone Promo, video 1 of 3. Swipe left or right to change the videos. Use the sound button to turn audio on or off. ## Move from a GWAS to a testable biological hypothesis Undertone turns a complete GWAS into ranked loci, fine-mapping, candidate genes, tissues and mechanism hypotheses. - Whole-study analysis and automatic locus discovery - Fine-mapping, colocalization and regulatory evidence - Private, reproducible reports with complete provenance ## A living map of scientific literature Noodle helps you discover biomedical research by meaning, explore connected publications and discuss scientific questions and papers with an active community of researchers. - Semantic discovery across biomedical literature - Discuss scientific questions and publications - Join an active community of researchers Paused until Noodle is in focus Noodle Promo, video 1 of 4. Swipe left or right to change the media. Use the sound button to turn audio on or off. Paused until Evidence is in focus Evidence Promo, video 1 of 3. Swipe left or right to change the videos. Use the sound button to turn audio on or off. ## Publish the study. Let readers test the decisions Evidence presents methods, results, sensitivity analyses and source files as one versioned study. Readers can change documented analytical decisions and see how the evidence moves. - Versioned study releases - Interactive sensitivity checks - GitHub-linked data and code ## We build computational methods, genomic infrastructure and clinical systems Helena brings together genomic methodology, high-performance computing and clinical software engineering. Each layer is developed as part of the same system, from the treatment of evidence and reference data to the way results reach laboratory review. ### Computational methods Deterministic classification, evidence handling, phenotype-driven prioritisation and methodologies for nuclear, mitochondrial and structural variation. ### Genomic infrastructure Locally indexed reference data, parallel processing, controlled environments and versioned analytical dependencies. ### Clinical systems Software that turns these methods into reviewable workflows for laboratories, geneticists and clinical teams. ## We develop the methods behind genomic interpretation Our work addresses how genomic evidence is admitted, classified, prioritised and presented for clinical review. Helena develops methodologies for deterministic variant classification, structural variation and agentic interpretation, and contributes these ideas through publications, preprints and technical documentation. Published · Frontiers in Genetics ### Real-world performance methodology Evaluation methods for concordance, change attribution and failure analysis against documented laboratory comparators. Early-stage variant classification platforms need a performance method that is stronger than an internal comparison but feasible before a multi-adjudicator regulatory study. A single peer laboratory is difficult to treat as ground truth when qualified laboratories can disagree. Formal adjudication, however, requires people and quality infrastructure that a small developer may not yet have. The paper proposes a three-layer performance model, an evidence-strength-ordered approach to ground truth, a six-category disposition taxonomy for disagreements and a phased route toward later regulatory evidence. Each observation is assigned to analytical, classification or clinical performance so that the source of a discrepancy remains visible. The method is demonstrated on 97 scored cases across three classifier versions. The demonstration shows how the framework operates. It does not validate the platform or establish superiority over another system. This is a methodological proposal with a single-platform demonstration. It uses one peer comparator, develops the taxonomy on the same cohorts used to illustrate it and carries conflicts of interest that require independent replication. Preprint · Perspective ### Evidence admissibility and deterministic classification A framework that keeps model output outside the class calculation and places human authorization where evidence enters the classifier. Clinical AI oversight is often concentrated at final output. In variant classification, an autonomous component can affect the decision earlier by changing the reference data and evidence snapshot consumed by the classifier. This Perspective places the load-bearing human transition at evidence admissibility (G1). An authorized person promotes a provenanced, version-locked evidence snapshot before it enters a deterministic, model-output-free ACMG/AMP classification core. Qualified sign-out (G4) remains necessary when the result is released. The paper proposes six evaluation metrics covering unauthorized evidence admission, attribution of classification changes, model-to-core boundary violations, exact reconstruction, narrative contradiction and prioritization omission. These are proposed tests, not reported validation results. The framework addresses auditability and governance. It does not establish classification correctness, clinical validation, comparative effectiveness or fitness for purpose. Complete separation also increases the workload of human literature curation. ## Work with us We at Helena Bioinformatics work with laboratories, universities and research organisations on genomic interpretation, validation and clinical implementation. ### Links and cited sources - [Create a free account](https://account.helena.bio/signup) - [Sign in](https://folklore.helena.bio/app/api/auth/login?return_to=%2Fapp) - [https://www.helena.bio/#company-folklore](https://www.helena.bio/#company-folklore) - [Explore Folklore](https://folklore.helena.bio/) - [https://www.helena.bio/#company-undertone](https://www.helena.bio/#company-undertone) - [Explore Undertone](https://undertone.helena.bio/) - [https://www.helena.bio/#company-noodle](https://www.helena.bio/#company-noodle) - [Explore Noodle](https://noodle.helena.bio/) - [https://www.helena.bio/#company-evidence](https://www.helena.bio/#company-evidence) - [Explore Evidence](https://evidence.helena.bio/) - [https://www.helena.bio/#company-capabilities](https://www.helena.bio/#company-capabilities) - [https://www.helena.bio/#work](https://www.helena.bio/#work) - [View research](https://www.helena.bio/research) - [https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2026.1925492/full](https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2026.1925492/full) - [https://zenodo.org/records/21346784](https://zenodo.org/records/21346784) - [https://www.helena.bio/#latest-news](https://www.helena.bio/#latest-news) - [Explore partnerships](https://www.helena.bio/partners) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Studies](https://folklore.helena.bio/studies) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Team | Helena Bioinformatics Source: https://www.helena.bio/team Canonical: https://www.helena.bio/team Description: Meet the people behind Helena Bioinformatics, bringing together genomics, clinical expertise and software engineering. # Meet the people behind Helena We bring together genomics, clinical expertise and software engineering. We work with laboratories, researchers and academic partners to turn scientific methods into systems used in real genomic work. ## Leadership ### Vladimir Mitev Founder & CEO Vladimir founded Helena Bioinformatics to bring scientific research, engineering and clinical genomics into practical systems for laboratories and partners. ### Academician Prof. Draga Toncheva, MD, DSc Founding Chair, Scientific & Clinical Advisory Board A pioneer of medical genetics and genomic medicine in Bulgaria, Academician Draga Toncheva brings more than four decades of scientific, clinical and academic experience. As Founding Chair, she advises Helena on scientific direction, clinical validation strategy, evidence boundaries and responsible human oversight. ## Scientific & Clinical Advisory Board ### Sena Karachanak-Yankova, PhD Associate Professor of Genetics | Independent Board Member Sena is an associate professor of genetics whose work covers medical genomics, rare and complex diseases, and human genetic variation. Her expertise supports the interpretation of whole-exome and whole-genome data, particularly where disease-associated variants must be understood in their clinical and population context. ### Svetoslav Dimov, PhD Associate Professor of Genetics | Independent Board Member Svetoslav is an associate professor of genetics whose work spans metagenomics, microbial genomics and molecular genetics. He brings specialist knowledge in the design and validation of metagenomic workflows, including microbial community analysis, bacterial genomics and antimicrobial resistance. ### Marija Dimitrova, MD Clinical Genetics Consultant | Independent Board Member Marija is a medical doctor working in human genetics and clinical genomics. She contributes a clinical perspective to the interpretation of genetic findings, with particular attention to patient context, evidence assessment and the responsible use of genomic information in clinical practice. ### Slavica Josifovska, PhD Associate Professor of Molecular Biology and Genomics | Independent Board Member Slavica is an associate professor at the Institute of Biology, Ss. Cyril and Methodius University in Skopje, where she works in the Laboratory for Molecular Biology and Genomics. Her research and teaching cover molecular biology, molecular genetics, genomics and genetic engineering. She advises Helena on molecular and genomic methods and academic collaboration in Southeast Europe. ### Sasho Panov, PhD Professor of Molecular Biology and Genomics | Independent Board Member Sasho is a professor of molecular biology and Head of the Department of Molecular Biology and Genomics at Ss. Cyril and Methodius University in Skopje. His work focuses on molecular diagnostics, cancer genomics and the genomic profiling of solid tumors, including the interpretation of clinically relevant molecular alterations. His contribution to the Board is centered on oncology genomics and the future development of somatic-variant analysis. ## Clinical validation ### Daniela Maneva Clinical Validation Contributor Daniela is a molecular geneticist with academic training in molecular biology, genetics, genomics and reproductive biology. She works with genomic and molecular evidence in a laboratory setting. She supported the clinical validation of ACMG classification rules by reviewing classifier outputs against case evidence and documenting rule behaviour that required correction or clarification. ### Spasimir Shishinyov Clinical Validation Contributor Spasimir is a molecular geneticist with experience in clinical genomics, rare-disease analysis and laboratory genetics. His work has included genomic research, clinical case analysis and the evaluation of molecular evidence. As a Clinical Validation Contributor, he reviewed Folklore's classifier outputs against case evidence and helped identify rules that required correction or clarification. ### Links and cited sources - [https://www.helena.bio/team#leadership](https://www.helena.bio/team#leadership) - [Vladimir Mitev](https://www.helena.bio/people/vladimir-mitev) - [https://www.linkedin.com/in/vlmitev/](https://www.linkedin.com/in/vlmitev/) - [https://www.linkedin.com/in/sena-karachanak-yankova-359539b3/](https://www.linkedin.com/in/sena-karachanak-yankova-359539b3/) - [https://www.linkedin.com/in/dimitrova-marija-8a6267145/](https://www.linkedin.com/in/dimitrova-marija-8a6267145/) - [https://www.linkedin.com/in/slavica-josifovska-b9ab5182/](https://www.linkedin.com/in/slavica-josifovska-b9ab5182/) - [https://www.linkedin.com/in/sasho-panov-9559777a/](https://www.linkedin.com/in/sasho-panov-9559777a/) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Vladimir Mitev, Founder and CEO | Helena Bioinformatics Source: https://www.helena.bio/people/vladimir-mitev Canonical: https://www.helena.bio/people/vladimir-mitev Description: Vladimir Mitev is the founder and CEO of Helena Bioinformatics and leads the development of Folklore, a platform for genomic variant interpretation. Founder profile # Vladimir Mitev Founder and CEO, Helena Bioinformatics Vladimir Mitev is the founder and CEO of Helena Bioinformatics, a bioinformatics company based in Sofia, Bulgaria. He leads the development of Folklore and the company's work across genomic interpretation, computational infrastructure and scientific collaboration. Current work ## Building infrastructure for genomic interpretation At Helena Bioinformatics, Vladimir leads product architecture, software engineering, infrastructure, company strategy and commercial deployment. His primary product work is Folklore, Helena Bioinformatics's proprietary platform for genomic variant interpretation and case-based genomic analysis. Folklore brings classification, phenotype, inheritance and literature evidence into one case record for specialist review. His work also includes reproducible genomics research, public computational studies and collaboration with laboratories, universities and scientific organisations. Areas of work Bioinformatics Genomic variant interpretation Clinical genomics software Cancer genomics Reproducible computational research Scientific and laboratory collaboration Selected public research ## Robustness and cross-cohort concordance of developmental regulons in endometrial carcinoma A public computational study repository containing the analysis record, reproducibility materials and project documentation. The study separates association, target-set robustness and transportability across cohorts. It evaluates developmental regulons in TCGA-UCEC and compares selected signals with CPTAC-UCEC. The public record explains which signals retained their direction and which remained unresolved, with limits on causal and clinical interpretation. Vladimir is listed as corresponding author alongside Draga Toncheva and Vasil Sgurev. The repository contains the computational analysis and provenance material; the research section of this website links the preprint and archived release. Readers can use those records to inspect the methods and evidence behind the summary, rather than treating the founder profile as a substitute for the study. The analysis asks whether developmental regulon signals in endometrial carcinoma survive changes in target composition and retain their direction in an independent cohort. It keeps pointwise association, target-set robustness and cross-cohort transportability as separate claims. Twenty prespecified developmental transcription-factor regulons and a fetal Mullerian epithelial module were evaluated in 507 TCGA-UCEC tumours. Six signals were then assessed in 230 CPTAC-UCEC tumours. The global fetal module was non-confirmatory. GATA2 and SOX9 retained the same direction across cohorts but failed the universal single-target-deletion gate. HOXA9 and WT1 lacked external confirmation. PAX8 and LHX1 were internally robust but remained externally unresolved. The results support directional concordance for selected regulons. They do not establish causal transcription-factor activity, biomarker validity, therapeutic dependence, treatment response or clinical utility. The preprint has not been peer reviewed. Profiles and identifiers ## GitHub Public research repositories and reproducibility records ## LinkedIn Professional profile and company activity ## Email vladimir@helena.bio ### Links and cited sources - [Helena Bioinformatics](https://www.helena.bio/) - [Explore Folklore](https://folklore.helena.bio/) - [View the research repository](https://github.com/helena-bioinformatics/endometrial-developmental-regulons) - [GitHubPublic research repositories and reproducibility records](https://github.com/vladimir-mitev) - [LinkedInProfessional profile and company activity](https://www.linkedin.com/in/vlmitev/) - [Emailvladimir@helena.bio](mailto:vladimir@helena.bio) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Research | Helena Bioinformatics Source: https://www.helena.bio/research Canonical: https://www.helena.bio/research Description: Research from Helena Bioinformatics on evidence admissibility, real-world performance and developmental regulons in endometrial carcinoma. # Advancing the science behind genomic technology We investigate the scientific questions that shape how genomic technologies are built, evaluated and used. We move the field forward through reproducible research, open methods and computational systems designed for real genomic work. The records below cover three different questions. Evidence admissibility examines who authorises the reference snapshot used by a classifier. The real-world performance framework separates analytical, classification and clinical performance and demonstrates a method on 97 scored cases. The endometrial carcinoma study tests developmental regulon associations, robustness and cross-cohort concordance. Each detail page includes the abstract, authors, publication date, version, DOI and evidence boundary. Follow the primary record to inspect the published article or preprint and, where available, its computational repository and archived release. A methodological proposal, a demonstration and an independently confirmed result carry different evidential weight. The pages preserve those distinctions and identify work that has not been peer reviewed. ## Research Published / Real-world performance ### A Methodological Framework for Real-World Performance Studies of Clinical Variant Classification Platforms at Early Organizational Stages A route between informal internal testing and formal regulatory adjudication, demonstrated on 97 scored cases. Early-stage variant classification platforms need a performance method that is stronger than an internal comparison but feasible before a multi-adjudicator regulatory study. A single peer laboratory is difficult to treat as ground truth when qualified laboratories can disagree. Formal adjudication, however, requires people and quality infrastructure that a small developer may not yet have. This is a methodological proposal with a single-platform demonstration. It uses one peer comparator, develops the taxonomy on the same cohorts used to illustrate it and carries conflicts of interest that require independent replication. 14 September 2026 Preprint / Clinical variant classification ### Evidence Admissibility Is the Load-Bearing Human Gate in Auditable Clinical Variant Classification Human authorization at evidence admissibility, before a version-locked snapshot enters deterministic classification. Clinical AI oversight is often concentrated at final output. In variant classification, an autonomous component can affect the decision earlier by changing the reference data and evidence snapshot consumed by the classifier. The framework addresses auditability and governance. It does not establish classification correctness, clinical validation, comparative effectiveness or fitness for purpose. Complete separation also increases the workload of human literature curation. 14 July 2026 Preprint / Precision oncology ### Robustness and cross-cohort concordance of developmental regulons in endometrial carcinoma Association, target-set robustness and transportability tested across TCGA-UCEC and CPTAC-UCEC. The analysis asks whether developmental regulon signals in endometrial carcinoma survive changes in target composition and retain their direction in an independent cohort. It keeps pointwise association, target-set robustness and cross-cohort transportability as separate claims. The results support directional concordance for selected regulons. They do not establish causal transcription-factor activity, biomarker validity, therapeutic dependence, treatment response or clinical utility. The preprint has not been peer reviewed. 2 August 2026 Find us on Read Vladimir Mitev's founder profile for his work and public research links. ### Links and cited sources - [Published / Real-world performanceA Methodological Framework for Real-World Performance Studies of Clinical Variant Classification Platforms at Early Organizational StagesA route between informal internal testing and formal regulatory adjudication, demonstrated on 97 scored cases.Early-stage variant classification platforms need a performance method that is stronger than an internal comparison but feasible before a multi-adjudicator regulatory study. A single peer laboratory is difficult to treat as ground truth when qualified laboratories can disagree. Formal adjudication, however, requires people and quality infrastructure that a small developer may not yet have.This is a methodological proposal with a single-platform demonstration. It uses one peer comparator, develops the taxonomy on the same cohorts used to illustrate it and carries conflicts of interest that require independent replication.14 September 2026Read more](https://www.helena.bio/research/real-world-performance) - [Preprint / Clinical variant classificationEvidence Admissibility Is the Load-Bearing Human Gate in Auditable Clinical Variant ClassificationHuman authorization at evidence admissibility, before a version-locked snapshot enters deterministic classification.Clinical AI oversight is often concentrated at final output. In variant classification, an autonomous component can affect the decision earlier by changing the reference data and evidence snapshot consumed by the classifier.The framework addresses auditability and governance. It does not establish classification correctness, clinical validation, comparative effectiveness or fitness for purpose. Complete separation also increases the workload of human literature curation.14 July 2026Read more](https://www.helena.bio/research/evidence-admissibility) - [Preprint / Precision oncologyRobustness and cross-cohort concordance of developmental regulons in endometrial carcinomaAssociation, target-set robustness and transportability tested across TCGA-UCEC and CPTAC-UCEC.The analysis asks whether developmental regulon signals in endometrial carcinoma survive changes in target composition and retain their direction in an independent cohort. It keeps pointwise association, target-set robustness and cross-cohort transportability as separate claims.The results support directional concordance for selected regulons. They do not establish causal transcription-factor activity, biomarker validity, therapeutic dependence, treatment response or clinical utility. The preprint has not been peer reviewed.2 August 2026Read more](https://www.helena.bio/research/endometrial-developmental-regulons) - [Vladimir Mitev's founder profile](https://www.helena.bio/people/vladimir-mitev) - [Vladimir Mitev and Helena Bioinformatics research on Zenodo](https://zenodo.org/search?q=metadata.creators.person_or_org.name%3A%22Mitev%2C+Vladimir%22) - [Vladimir Mitev on ORCID](https://orcid.org/0009-0001-8013-6850) - [Helena Bioinformatics on bio.tools](https://bio.tools/Folklore) - [Vladimir Mitev on Web of Science](https://www.webofscience.com/wos/author/record/QUT-3684-2026) - [Helena Bioinformatics Validation Studies on OSF](https://osf.io/xnfzd/) - [Vladimir Mitev on ResearchGate](https://www.researchgate.net/profile/Vladimir-Mitev-3) - [Vladimir Mitev on Google Scholar](https://scholar.google.com/citations?user=cdOyotsAAAAJ) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Evidence Admissibility in Variant Classification | Helena Source: https://www.helena.bio/research/evidence-admissibility Canonical: https://www.helena.bio/research/evidence-admissibility Description: Human authorization at evidence admissibility, before a version-locked snapshot enters deterministic classification. # Evidence Admissibility Is the Load-Bearing Human Gate in Auditable Clinical Variant Classification Vladimir Mitev Published 14 July 2026 Record type Preprint Version Version 2.0 Article DOI 10.5281/zenodo.21346784 Abstract Clinical AI oversight is often concentrated at final output. In variant classification, an autonomous component can affect the decision earlier by changing the reference data and evidence snapshot consumed by the classifier. This Perspective places the load-bearing human transition at evidence admissibility (G1). An authorized person promotes a provenanced, version-locked evidence snapshot before it enters a deterministic, model-output-free ACMG/AMP classification core. Qualified sign-out (G4) remains necessary when the result is released. The paper proposes six evaluation metrics covering unauthorized evidence admission, attribution of classification changes, model-to-core boundary violations, exact reconstruction, narrative contradiction and prioritization omission. These are proposed tests, not reported validation results. Evidence boundary The framework addresses auditability and governance. It does not establish classification correctness, clinical validation, comparative effectiveness or fitness for purpose. Complete separation also increases the workload of human literature curation. 01 ## Where the human gate sits G1 governs which evidence snapshot may enter the classifier. The classification core then produces a frozen class and criterion trace without model output. Language-model agents remain downstream, where they can support interpretation without changing the computed class. G4 governs qualified release. The two gates protect different transitions: evidence admission before computation and sign-out after computation. 02 ## How the proposal can be tested The proposed metrics ask whether evidence changes were authorized, whether classification drift can be attributed to a recorded change, whether model-generated material crossed into the class-determining path and whether a result can be reconstructed from its manifest. The revised record reports no new dataset or completed validation study. It defines the architecture, the gates and the measurements needed to test them. Authors and affiliations Vladimir MitevHelena Bioinformatics, Sofia, Bulgaria Vladimir Mitev Helena Bioinformatics, Sofia, Bulgaria ## More research ### A Methodological Framework for Real-World Performance Studies of Clinical Variant Classification Platforms at Early Organizational Stages A route between informal internal testing and formal regulatory adjudication, demonstrated on 97 scored cases. Published / 14 September 2026 ### Robustness and cross-cohort concordance of developmental regulons in endometrial carcinoma Association, target-set robustness and transportability tested across TCGA-UCEC and CPTAC-UCEC. Preprint / 2 August 2026 ### Links and cited sources - [All research](https://www.helena.bio/research) - [10.5281/zenodo.21346784](https://doi.org/10.5281/zenodo.21346784) - [Open the Zenodo record](https://zenodo.org/records/21346784) - [A Methodological Framework for Real-World Performance Studies of Clinical Variant Classification Platforms at Early Organizational Stages](https://www.helena.bio/research/real-world-performance) - [Robustness and cross-cohort concordance of developmental regulons in endometrial carcinoma](https://www.helena.bio/research/endometrial-developmental-regulons) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Evaluating Variant Classification Platforms | Helena Source: https://www.helena.bio/research/real-world-performance Canonical: https://www.helena.bio/research/real-world-performance Description: A route between informal internal testing and formal regulatory adjudication, demonstrated on 97 scored cases. # A Methodological Framework for Real-World Performance Studies of Clinical Variant Classification Platforms at Early Organizational Stages Vladimir Mitev Frontiers in Genetics · Hypothesis and Theory. The manuscript has not been peer reviewed. Published 14 September 2026 Record type Published Version Current public record Article DOI 10.3389/fgene.2026.1925492 Abstract Early-stage variant classification platforms need a performance method that is stronger than an internal comparison but feasible before a multi-adjudicator regulatory study. A single peer laboratory is difficult to treat as ground truth when qualified laboratories can disagree. Formal adjudication, however, requires people and quality infrastructure that a small developer may not yet have. The paper proposes a three-layer performance model, an evidence-strength-ordered approach to ground truth, a six-category disposition taxonomy for disagreements and a phased route toward later regulatory evidence. Each observation is assigned to analytical, classification or clinical performance so that the source of a discrepancy remains visible. The method is demonstrated on 97 scored cases across three classifier versions. The demonstration shows how the framework operates. It does not validate the platform or establish superiority over another system. Evidence boundary This is a methodological proposal with a single-platform demonstration. It uses one peer comparator, develops the taxonomy on the same cohorts used to illustrate it and carries conflicts of interest that require independent replication. 01 ## Four parts of the framework The performance model separates analytical, classification and clinical questions. Ground truth is assembled from more than one source and weighted by evidence strength. Disagreements are assigned to six methodological dispositions with different corrective actions. A phased pathway preserves early evidence for later, more formal evaluation. 02 ## What the demonstration contributes The 97 scored cases show how a team can record discrepancies, attribute them to the correct performance layer and avoid treating every platform-comparator difference as a classifier defect. The paper claims a method. Validation would require independent cohorts, broader comparators and adjudication that is separate from the team that developed the framework. Authors and affiliations Vladimir MitevHelena Bioinformatics, Sofia, Bulgaria Vladimir Mitev Helena Bioinformatics, Sofia, Bulgaria ## More research ### Evidence Admissibility Is the Load-Bearing Human Gate in Auditable Clinical Variant Classification Human authorization at evidence admissibility, before a version-locked snapshot enters deterministic classification. Preprint / 14 July 2026 ### Robustness and cross-cohort concordance of developmental regulons in endometrial carcinoma Association, target-set robustness and transportability tested across TCGA-UCEC and CPTAC-UCEC. Preprint / 2 August 2026 ### Links and cited sources - [All research](https://www.helena.bio/research) - [10.3389/fgene.2026.1925492](https://doi.org/10.3389/fgene.2026.1925492) - [Read in Frontiers](https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2026.1925492/full) - [Open the Zenodo record](https://zenodo.org/records/21105027) - [Evidence Admissibility Is the Load-Bearing Human Gate in Auditable Clinical Variant Classification](https://www.helena.bio/research/evidence-admissibility) - [Robustness and cross-cohort concordance of developmental regulons in endometrial carcinoma](https://www.helena.bio/research/endometrial-developmental-regulons) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Endometrial Carcinoma Regulon Study | Helena Source: https://www.helena.bio/research/endometrial-developmental-regulons Canonical: https://www.helena.bio/research/endometrial-developmental-regulons Description: Association, target-set robustness and transportability tested across TCGA-UCEC and CPTAC-UCEC. # Robustness and cross-cohort concordance of developmental regulons in endometrial carcinoma Draga Toncheva, Vasil Sgurev, Vladimir Mitev Submitted to npj Precision Oncology. The manuscript has not been peer reviewed. Published 2 August 2026 Record type Preprint Version Version 1.0 Article DOI 10.5281/zenodo.21763096 Abstract The analysis asks whether developmental regulon signals in endometrial carcinoma survive changes in target composition and retain their direction in an independent cohort. It keeps pointwise association, target-set robustness and cross-cohort transportability as separate claims. Twenty prespecified developmental transcription-factor regulons and a fetal Mullerian epithelial module were evaluated in 507 TCGA-UCEC tumours. Six signals were then assessed in 230 CPTAC-UCEC tumours. The global fetal module was non-confirmatory. GATA2 and SOX9 retained the same direction across cohorts but failed the universal single-target-deletion gate. HOXA9 and WT1 lacked external confirmation. PAX8 and LHX1 were internally robust but remained externally unresolved. Evidence boundary The results support directional concordance for selected regulons. They do not establish causal transcription-factor activity, biomarker validity, therapeutic dependence, treatment response or clinical utility. The preprint has not been peer reviewed. 01 ## Analysis design TCGA-UCEC served as the discovery cohort. The analysis tested molecular-class associations and then removed individual targets to see whether a regulon-level result depended on one gene. CPTAC-UCEC provided an external check on the direction of six selected signals. This sequence prevents an association from being described as robust or transportable before it passes the corresponding test. 02 ## Reproducibility record The public repository contains the TCGA-UCEC discovery analysis, CPTAC-UCEC external evaluation, sensitivity analyses, provenance records, machine-readable results and release-verification tooling. Repository concept DOI 10.5281/zenodo.21762178 identifies the computational record. Immutable release DOI 10.5281/zenodo.21762179 identifies version v1.0.1. Authors and affiliations Draga TonchevaDepartment of Biological Sciences, Bulgarian Academy of Sciences, Sofia, BulgariaVasil SgurevInstitute of Information and Communication Technologies, Bulgarian Academy of Sciences, Sofia, BulgariaVladimir Mitev (corresponding)Helena Bioinformatics EOOD, Sofia, Bulgaria Draga Toncheva Department of Biological Sciences, Bulgarian Academy of Sciences, Sofia, Bulgaria Vasil Sgurev Institute of Information and Communication Technologies, Bulgarian Academy of Sciences, Sofia, Bulgaria Vladimir Mitev (corresponding) Helena Bioinformatics EOOD, Sofia, Bulgaria Correspondence Vladimir Mitevvladimir@helena.bio Repository DOI 10.5281/zenodo.21762178 Immutable release DOI 10.5281/zenodo.21762179 ## More research ### A Methodological Framework for Real-World Performance Studies of Clinical Variant Classification Platforms at Early Organizational Stages A route between informal internal testing and formal regulatory adjudication, demonstrated on 97 scored cases. Published / 14 September 2026 ### Evidence Admissibility Is the Load-Bearing Human Gate in Auditable Clinical Variant Classification Human authorization at evidence admissibility, before a version-locked snapshot enters deterministic classification. Preprint / 14 July 2026 ### Links and cited sources - [All research](https://www.helena.bio/research) - [10.5281/zenodo.21763096](https://doi.org/10.5281/zenodo.21763096) - [vladimir@helena.bio](mailto:vladimir@helena.bio) - [10.5281/zenodo.21762178](https://doi.org/10.5281/zenodo.21762178) - [10.5281/zenodo.21762179](https://doi.org/10.5281/zenodo.21762179) - [View the research repository](https://github.com/helena-bioinformatics/endometrial-developmental-regulons) - [A Methodological Framework for Real-World Performance Studies of Clinical Variant Classification Platforms at Early Organizational Stages](https://www.helena.bio/research/real-world-performance) - [Evidence Admissibility Is the Load-Bearing Human Gate in Auditable Clinical Variant Classification](https://www.helena.bio/research/evidence-admissibility) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Careers at Helena Bioinformatics | Helena Bioinformatics Source: https://www.helena.bio/careers Canonical: https://www.helena.bio/careers Description: Meet Helena Bioinformatics and introduce yourself for future opportunities in clinical genomics, bioinformatics and scientific software. Careers # Work with Helena We do not have open positions at the moment, but we are always interested in meeting thoughtful people who want to work on clinical genomics, bioinformatics and scientific software. Tell us what you work on, your interests and where you think you could contribute. Stay in touch ## Join our talent network There are no open positions right now, but the right opportunity may open tomorrow. Introduce yourself and stay on our radar. Your information is kept confidential and used only to contact you about future opportunities. ## Introducing your work Our work connects software engineering with the evidence that specialists use to interpret genomic variants. Folklore brings classification, phenotype, inheritance and literature into a case record for review. Other Helena projects address genetic association analysis, biomedical literature discovery and reproducible public study records. The company website and public research pages provide examples of these different areas. A useful introduction describes a problem you have worked on, your contribution and how you checked the result. You can include links to a public repository, publication, portfolio or technical explanation. Tell us which part of clinical genomics, bioinformatics or scientific software interests you and how it relates to your experience. Share only material you are entitled to make available; a public example or a description is enough. This page is an invitation to introduce yourself for future opportunities, rather than an advertised vacancy. We have no open positions at present. Use the contact button above to explain your interests and provide an email address where we can reach you. An engineering example can explain the inputs, the result and the checks that make the result reproducible. A public code sample is easier to understand when it includes a short description of the problem, the part you implemented and the conditions under which it was tested. If the original work is confidential, describe your contribution without attaching private source code, credentials or customer data. For a scientific example, identify the question, the data source, the method and the limits of the conclusion. Our public research distinguishes an association from a robust result and an independently confirmed finding. A publication, a reproducible analysis or a clear account of an unresolved result can help explain how you approach that distinction in your own work. An introduction can also describe collaboration across disciplines: for example, how you explained a computational result to a laboratory colleague or translated a scientific requirement into software behaviour. State your own role and the part of the work you would like to discuss. These suggestions help give an introduction context; they are not requirements for an advertised position. ### Links and cited sources - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Contact | Helena Bioinformatics Source: https://www.helena.bio/contact Canonical: https://www.helena.bio/contact Description: Get in touch with Helena Bioinformatics - partnerships, demos, and general inquiries. Contact # Contact Us Contact Helena about a laboratory workflow, research collaboration, product demonstration, or general question. ## General Inquiries Questions about Helena, pricing, or partnerships. ## Company Helena Bioinformatics Sofia, Bulgaria ## Data Protection GDPR, DPA, and privacy-related inquiries. ## Before you send a message For a product discussion, tell us which Helena product you are asking about and what you would like to do. A laboratory evaluating Folklore can describe its current interpretation workflow, the type of input it works with and the questions it wants a demonstration to address. General questions about pricing, partnerships and the company can go to contact@helena.bio. For research collaboration, include the scientific question and links to relevant public material. Our research pages distinguish published preprints, computational records and the limits of the evidence. Referring to a specific paper, study or product page helps us understand the context of your enquiry without requiring you to reproduce the entire background. The message form asks for your name, email, organisation, topic and message. Choose the subject that best describes your enquiry and include a clear question or next step. For a technical question, name the product and describe the behaviour you encountered. Please keep the initial message free of patient information, genomic files, passwords and other confidential records. Questions about personal-data processing, the Data Processing Agreement or the Privacy Notice belong with privacy@helena.bio. The linked Privacy Policy explains how information submitted through this contact route is handled. Helena Bioinformatics is based in Sofia, Bulgaria; the email addresses on this page provide direct routes for general and data-protection enquiries. ### Questions about a product Name the product and the workflow you want to discuss. A question about interpreting a genomic case belongs to a different conversation from a question about literature discovery or a public study record. Include the public page you have read and the point that remains unclear. For a demonstration request, describe what you would like to see rather than attaching a real case. ### Questions about published research Include the paper title or DOI and, where relevant, the version you are reading. A question about a method, an interpretation of a result and a request for reproducibility material may need different source records. Describe the passage, figure or stated limitation that prompted your question so the conversation can start from the same evidence. ### Technical questions Describe the action, the expected behaviour and what happened instead. A page URL and a short sequence of steps are useful context. Remove personal information from any description and keep access tokens, passwords, genomic files and medical reports out of the initial enquiry. The public form is intended to start a conversation, not to transfer clinical data. By contacting us, your data will be processed in accordance with our Privacy Policy. ### Links and cited sources - [contact@helena.bio](mailto:contact@helena.bio) - [privacy@helena.bio](mailto:privacy@helena.bio) - [Privacy Policy](https://www.helena.bio/privacy) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Data Processing Agreement | Helena Bioinformatics Source: https://www.helena.bio/dpa Canonical: https://www.helena.bio/dpa Description: Article 28 GDPR terms for customer-controlled personal data. # Data Processing Agreement Version 1.0 · Effective: 14 August 2026 This Data Processing Agreement (“DPA”) forms part of the agreement under which Helena Bioinformatics EOOD provides the Service to the customer identified in an order, registration record or other written agreement (“Controller”). Helena is the “Processor”. This DPA applies only where Helena processes personal data on behalf of the Controller. It must be accepted or signed by a person authorized to bind the Controller. It does not apply merely because a person creates an Individual or Distributor profile. ## 1. Definitions and interpretation “Controller Data” means personal data processed by Helena on behalf of the Controller under the Service. “Data Protection Law” means the GDPR, applicable Bulgarian data-protection law and other binding law applicable to the processing. “Subprocessor”, “personal data breach”, “processing”, “controller”, “processor” and “data subject” have the meanings in Data Protection Law. If this DPA conflicts with the main service agreement on data-protection matters, this DPA prevails. An order may specify additional safeguards but may not reduce mandatory protections. ## 2. Details and duration of processing The subject matter, nature, purpose, data categories, data subjects and processing operations are described in Annex 1. Processing continues for the term of the Service and any limited return, deletion, backup-expiry or legal-retention period. The Controller instructs Helena to process Controller Data as necessary to provide, secure, support and maintain the contracted Service; follow documented product configuration and support requests; and comply with this DPA and applicable law. ## 3. Controller obligations The Controller determines the purposes and means of processing and is responsible for lawful instructions, transparency, Article 6 and Article 9 legal bases, data-subject rights, data accuracy and minimization. It must ensure that persons submitting data are authorized and that direct identifiers are not uploaded unless expressly agreed. The Controller must not instruct Helena to process data unlawfully or outside the agreed Service. It must maintain reasonable account security, assign appropriate roles and notify Helena of relevant restrictions, requests and incidents without undue delay. ## 4. Processor obligations and instructions Helena will process Controller Data only on documented instructions, including for international transfers, unless Union or Member State law requires processing. In that case Helena will inform the Controller before processing unless the law prohibits notice on important public-interest grounds. Helena will immediately inform the Controller if, in its opinion, an instruction infringes Data Protection Law and may suspend the affected processing until the instruction is amended or confirmed. Helena will ensure authorized personnel are bound by confidentiality and access data only as necessary. ## 5. Security of processing Helena will implement and maintain appropriate technical and organizational measures taking account of the state of the art, implementation cost, scope, context and risk, including the risk presented by genetic and health data. Current measures are described in Annex 2. Helena may update measures as technology and risk evolve, provided the overall level of protection is not materially reduced. The Controller acknowledges that security is a shared responsibility and will use available configuration and access controls appropriately. ## 6. Subprocessors The Controller gives general written authorization for the subprocessors listed in Annex 3. Helena will impose data-protection obligations providing substantially equivalent protection and remains responsible for their performance to the extent required by Article 28 GDPR. Helena will give at least 30 days’ prior notice of an intended addition or replacement that will process Controller Data. The Controller may object on reasonable documented data-protection grounds. The parties will work in good faith on a reasonable alternative; if none is available, either party may terminate only the affected Service without penalty for the unused affected period. A provider used only for public websites, ordinary business administration or payment processing is not a subprocessor under this DPA unless it receives Controller Data. ## 7. International transfers Helena will host private genomic processing in the EEA unless the order expressly states otherwise. Helena will not transfer Controller Data outside the EEA without documented Controller authorization and a lawful Chapter V GDPR mechanism. Where required, the parties incorporate the applicable European Commission Standard Contractual Clauses and supplementary measures. Helena will provide information reasonably required for the Controller’s transfer assessment. ## 8. Data-subject requests Taking account of the nature of processing, Helena will provide reasonable technical and organizational assistance for the Controller to respond to requests under Chapter III GDPR. Helena will not respond substantively on the Controller’s behalf unless instructed or legally required. If Helena receives a request relating to Controller Data, it will forward it to the Controller without undue delay where the requester and Controller can reasonably be identified. ## 9. Compliance assistance Taking account of the nature of processing and information available, Helena will assist the Controller with Articles 32–36 GDPR, including security, breach assessment, data-protection impact assessments and prior consultation. Extraordinary assistance outside normal Service operation may be charged at an agreed reasonable rate unless caused by Helena’s breach. ## 10. Personal-data breaches Helena will notify the Controller without undue delay after becoming aware of a personal-data breach affecting Controller Data and, where practicable, within 24 hours. The notice will provide available information about the nature, likely consequences, affected categories and approximate numbers, contact point and mitigation. Information may be supplied in phases. Notification is not an admission of fault. The Controller is responsible for notifications to supervisory authorities and data subjects, and Helena will reasonably assist. ## 11. Audit and information rights Helena will make available information reasonably necessary to demonstrate Article 28 compliance. The Controller should first use current security documentation, certifications and independent reports made available by Helena. If that information is insufficient, the Controller may conduct one audit per year on at least 30 days’ notice, during normal business hours, through a qualified independent auditor bound by confidentiality. Additional audits are permitted after a material breach or where a supervisory authority requires them. Audits must avoid access to other customers’ data, disruption and security compromise. ## 12. Return and deletion At the Controller’s choice, Helena will return or delete Controller Data after the Service ends, unless law requires retention. The Controller must communicate its choice before termination or within the period stated in the order; otherwise Helena may delete according to the documented retention schedule. Deletion applies to active systems and derived analysis results in scope. Backup copies are isolated from ordinary use and expire through the backup lifecycle unless restoration is required for disaster recovery, in which case deletion restrictions are reapplied. Helena will provide reasonable deletion confirmation on request. ## 13. Government and third-party requests Unless prohibited by law, Helena will notify the Controller of a binding request for Controller Data and will challenge requests it reasonably considers unlawful or disproportionate. Helena will disclose only the minimum legally required data. ## 14. Liability and term This DPA begins when the main agreement becomes effective or an authorized representative accepts it and remains in force while Helena processes Controller Data. Liability is governed by the main agreement, subject to liabilities and data-subject rights that cannot be limited under Data Protection Law. ## 15. Governing law and changes This DPA is governed by Bulgarian law and directly applicable European Union law. Competent courts in Sofia have jurisdiction unless mandatory law provides otherwise. Helena may update this standard DPA to reflect law or Service changes. A material reduction of protection requires notice and, where required, renewed agreement. Each version is preserved by version, language, effective date and content digest. ## Annex 1 — Processing details Subject matter and purpose: provision of genomic and biomedical analysis, evidence retrieval, prioritization, reporting, collaboration, support, security and related contracted functions. Operations: receipt, validation, storage, organization, parsing, annotation, comparison, classification support, querying, retrieval, generation, display, transmission to authorized users, backup, security logging, return and deletion. Data subjects: patients and relatives whose data is submitted by the Controller; authorized users; and other persons included in Controller Data under documented instructions. Data categories: pseudonymized sample identifiers; genetic variant and sequence-derived data; phenotype and health information; family relationships; clinical annotations; analysis parameters and results; reports; provenance; and access metadata. Direct identifiers are excluded unless specifically agreed. Special categories: genetic and health data under Article 9 GDPR. Frequency: as initiated by authorized users or configured workflows. Duration: the Service term plus the return, deletion, backup-expiry and legally required periods. ## Annex 2 — Technical and organizational measures Governance and people: documented access responsibilities, confidentiality obligations, least privilege, administrator controls, security and privacy review, incident response and change management. Identity and access: individual accounts, password hashing, role-based authorization, expiring sessions or tokens, privileged-access restrictions and auditable administrative actions. Multi-factor authentication is applied where configured and required by the applicable access policy. Infrastructure and communications: EEA-based private processing infrastructure, network filtering and service isolation, encrypted transport using current secure protocols, managed secrets and restricted administrative access. Data protection: pseudonymization requirements, minimization, separation between public and private AI paths, controlled storage, backup and recovery processes, retention configuration and deletion procedures. Encryption at rest is applied where supported and configured for the relevant storage layer; Helena does not represent that a single algorithm protects every medium. Assurance and continuity: security logging, monitoring, vulnerability and dependency management, tested deployment and migration procedures, backup validation, incident escalation and periodic control review. Exact retention periods and recovery objectives are those documented for the contracted configuration. ## Annex 3 — Authorized subprocessors Hetzner Online GmbH, Germany / Finland infrastructure region — dedicated server, network and storage infrastructure for private Service processing in the EEA. No public website, analytics, ordinary email, payment or public-AI provider is authorized by this Annex to receive private Controller Data. If a contracted feature requires another provider, Helena must add it through the notice and authorization process in Section 6 before that provider receives Controller Data. ## 16. Execution and contact Electronic acceptance by the Controller’s authorized laboratory representative, tied to the exact version and digest in Helena’s legal registry, constitutes written execution where permitted by law. Helena records the accepting user, organization, time, source and document digest. Data-protection notices under this DPA should be sent to privacy@helena.bio. Operational security incidents should also be sent to security@helena.bio. ### Links and cited sources - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Data Protection Impact Assessment | Helena Bioinformatics Source: https://www.helena.bio/dpia Canonical: https://www.helena.bio/dpia Description: DPIA summary for the Helena genomic variant analysis platform. # Data Protection Impact Assessment Last updated: March 2026 (v1.1) This document provides a summary of the Data Protection Impact Assessment (DPIA) conducted by Helena Bioinformatics for the Helena platform, pursuant to GDPR Article 35. A DPIA is mandatory when processing genetic data on a large scale, as it constitutes high-risk processing of special category data. ## 1. Processing Description Nature of processing: automated analysis of genetic variant data (VCF files) uploaded by clinical genetics laboratories. Processing includes variant annotation against population and clinical databases, automated ACMG/AMP classification, phenotype-genotype correlation, biomedical literature mining, AI-assisted clinical summarization, and generation of clinical interpretation reports. All AI models run locally on dedicated infrastructure and no genomic data is sent to external AI services or APIs. Scope: the platform processes whole-exome and whole-genome sequencing data containing thousands to millions of genetic variants per patient sample. Associated phenotype data (HPO terms) and clinical context are also processed. Context: the platform serves as a clinical decision support tool for qualified geneticists. It does not make autonomous clinical decisions. All outputs, including AI-generated summaries, require professional review and validation. Purpose: to reduce variant interpretation time from days to minutes, improving laboratory throughput and consistency while maintaining clinical accuracy. ## 2. Necessity and Proportionality Necessity: processing genetic variant data is essential to the core function of the platform. The service cannot be provided without processing VCF files. Phenotype data is necessary for clinical correlation and prioritization of variants. Proportionality: we process only the minimum data necessary. VCF files are received in pseudonymized form (sample IDs only, no patient names). We do not request or store directly identifying patient information. Phenotype data is limited to standardized HPO codes relevant to the clinical question. Data retention is time-limited with automatic deletion. Legal basis: Helena Bioinformatics operates in a dual capacity depending on the processing activity. When processing genomic data on behalf of a laboratory for clinical variant interpretation, Helena acts as Data Processor under the Data Processing Agreement (Article 28 GDPR). The Data Controller (laboratory) relies on explicit consent (Article 9(2)(a)) or the healthcare provision exemption (Article 9(2)(h)) as the legal basis for processing special category data. When processing de-identified data for internal research and development purposes, including algorithm validation and platform improvement, Helena acts as an independent Data Controller with legitimate interest (Article 6(1)(f)) as the legal basis, in accordance with the applicable Data Use Agreement. In all cases, the stricter conditions of the governing agreements apply. ## 3. Risk Assessment Risk 1: Unauthorized access to genetic data Severity: High. Genetic data is immutable and uniquely identifying. Likelihood: Low. Mitigated by dedicated servers (not multi-tenant cloud), TLS 1.3 encryption in transit, AES-128 symmetric encryption at application level (Fernet with HMAC-SHA256 integrity verification), full-disk encryption at infrastructure level, role-based access control, network firewall rules, and comprehensive audit logging. Residual risk: Low. Risk 2: Data breach during transit Severity: High. Likelihood: Very Low. All data transmission uses TLS 1.3 encryption. VCF files are uploaded directly to our EU servers. No genomic data transits through non-EU jurisdictions. Residual risk: Very Low. Risk 3: Re-identification of pseudonymized data Severity: High. Genetic data is inherently identifying. Likelihood: Very Low. We receive only sample IDs, not patient identifiers. Our staff cannot link sample IDs to individuals. Access is restricted to automated processing pipelines. Residual risk: Very Low. Risk 4: Incorrect variant classification leading to clinical harm Severity: High. Incorrect classification could affect patient care. Likelihood: Low. The platform is explicitly positioned as a decision support tool, not a diagnostic device. All outputs must be reviewed by qualified geneticists. The platform follows established ACMG/AMP guidelines and references validated databases (ClinVar, gnomAD). Regular validation against clinical benchmarks. Residual risk: Low (mitigated by mandatory human review). Risk 5: Excessive data retention Severity: Medium. Likelihood: Very Low. Automatic deletion after configurable retention period (default 90 days). Data Controllers can request immediate deletion. Deletion processes are logged and auditable. Residual risk: Very Low. Risk 6: Sub-processor non-compliance Severity: Medium. Likelihood: Very Low. Hetzner Online GmbH (infrastructure provider) maintains ISO 27001 certification and provides physical hosting only without logical data access. DPA in place with Hetzner. Vercel Inc. (frontend hosting) processes only website delivery assets and user interface resources; no genomic data, patient identifiers, or clinical information is processed by or accessible to Vercel. Standard Contractual Clauses apply for any data transfers outside the EU. No sub-processors outside the EU process genomic data. Residual risk: Very Low. Risk 7: AI model misuse or data leakage through AI processing Severity: High. Likelihood: Very Low. All AI and machine learning models operate exclusively on dedicated local GPU infrastructure owned and managed by Helena Bioinformatics within the same EU datacenter. No genomic or clinical data is transmitted to external AI services, cloud-based language models, or third-party APIs. AI processing is subject to the same access controls, audit logging, and encryption as all other platform operations. Residual risk: Very Low. ## 4. Measures to Address Risks The following technical and organizational measures are implemented: Technical measures: dedicated EU-based servers (Helsinki, Finland) with no multi-tenant sharing; TLS 1.3 encryption in transit; AES-128 symmetric encryption at application level with HMAC-SHA256 integrity verification, plus full-disk encryption at infrastructure level; role-based access control with principle of least privilege and organization-level data isolation; JWT authentication with automatic session expiration; comprehensive audit trails for all data operations; automated data deletion based on configurable retention policies; network isolation with firewall restricting non-essential traffic; bcrypt password hashing (work factor 12); API key management with SHA-256 hashing and constant-time comparison; local-only AI inference with no external API calls; and regular vulnerability assessments. Organizational measures: Data Processing Agreements with all laboratory partners; confidentiality obligations for all personnel; documented security incident response procedures with 24-hour notification to Data Controllers; regular review and updating of security measures; sub-processor management with Data Controller notification; staff training on data protection obligations; and designated data protection contact. ## 5. Regulatory Compliance The Helena platform is designed and operated in compliance with applicable EU and national data protection and cybersecurity regulations. In addition to GDPR (Regulation (EU) 2016/679), the platform's security architecture addresses the requirements of the NIS2 Directive (Directive (EU) 2022/2555) as transposed into Bulgarian national law through the amendments to the Cybersecurity Act (State Gazette No. 17, 13 February 2026). While Helena Bioinformatics may not itself be classified as an essential or important entity under NIS2, its infrastructure and security measures are designed to support laboratory partners who hold such status. The measures described in this DPIA, including encryption at rest and in transit, access control, audit logging, incident response procedures, and sub-processor management, are aligned with the risk management obligations set out in NIS2 Article 21. ## 6. Conclusion This DPIA concludes that the Helena platform can process genetic data with an acceptable level of residual risk, provided all identified measures are maintained and regularly reviewed. The key factors supporting this conclusion are: data is received only in pseudonymized form; all processing occurs within the EU on dedicated infrastructure; encryption is applied both in transit and at rest; AI models operate exclusively on local infrastructure with no external data transmission; the platform functions as decision support requiring mandatory human review; and data retention is time-limited with automatic deletion. This DPIA will be reviewed annually or when significant changes are made to the processing activities, infrastructure, or regulatory landscape. For questions regarding this assessment, contact us at privacy@helena.bio. ## Version History v1.1 - March 2026: clarified dual legal basis (Processor and Controller roles); corrected encryption specification (AES-128 Fernet at application level, full-disk encryption at infrastructure level); added Risk 7 covering AI processing and local-only inference; added Vercel as sub-processor with scope clarification; added Section 5 on NIS2 regulatory compliance; updated DPO reference to designated data protection contact; general editorial improvements. v1.0 - February 2026: initial publication. ### Links and cited sources - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Investors | Helena Bioinformatics Source: https://www.helena.bio/investors Canonical: https://www.helena.bio/investors Description: Company, evidence, market and financing information for investors evaluating Helena Bioinformatics and its Folklore genomic interpretation platform. # Genomic interpretation software for European laboratories We build FOLKLORE for medical genetics laboratories. It processes panel, exome and genome VCFs and brings classification, phenotype, inheritance and literature evidence into one record for review by a qualified specialist. Sofia · Bulgaria Seed · TRL 6 Europe-first ## Clinical validation 197 ### Clinical testing Retrospective cases across four documented whole-genome cohorts. Operational ### Current status FOLKLORE is in operational use within a laboratory workflow. <10 min ### WGS processing time Classification and prioritisation on dedicated infrastructure. TRL 6 ### Assessment scope TRL 6 for the nuclear SNV/indel core and specialist-review workflow. 157 ### Locked-version cases Processed under classifier version v3.39.1. EU ### Processing location Genomic processing and model-assisted interpretation run on EU infrastructure. ### Cross-cohort studies The programme now spans 197 cases across four cohorts. Cohort 4 completed 100/100 cases, produced 67.0% FULL plus CLINICAL agreement, and converted a bounded forensic defect set into a defined correction and full-cohort regression programme. The runtime covers computational classification and prioritisation of a whole-genome VCF on Helena's dedicated infrastructure. Specialist review follows that processing stage. ## Sequencing capacity has grown faster than interpretation capacity ### One case record FOLKLORE stores annotation, deterministic classification, phenotype, inheritance and literature evidence in the same case record used for specialist review. ### Per-case access A laboratory can begin with its own cases, then purchase individual cases, prepaid volumes or an annual deployment. ### EU-based processing Primary genomic processing runs on European infrastructure. A qualified specialist reviews the evidence and authorises the final interpretation and report. ## Market focus We are starting with laboratories and rare-disease teams that process approximately 50-500 inherited-disease WES or WGS cases each year. Hospital genomics units processing up to 3,000 cases are the next customer group. Commercial path Own-data pilot → pay per case → prepaid volume packs → annual commitment or private instance ### European serviceable market 500-1,000 qualified EU/EEA accounts, based on internal bottom-up planning assumptions. €4.5m–€40m ### Five-year base case 50 paying accounts processing an average of 250 cases. €1.0m ### Five-year growth case 125 accounts plus a limited number of enterprise and API contracts. ~€5.1m We calculate these ranges from account counts, expected annual case volume and net revenue per case. We will qualify and deduplicate the European laboratory list country by country. ## Laboratory use, distribution and research agreements We work with CellGenetics in the current laboratory environment and with ELTA 90 through our signed Bulgarian reseller agreement. Our research memoranda cover genomic data integration, laboratory systems and academic work. ### CellGenetics Active clinical-validation relationship and operational RUO use in a real laboratory workflow. ### ELTA 90 Signed exclusive reseller for the Bulgarian channel, with contracted pricing, onboarding and support roles. ### Research network Cooperation with IICT-BAS, the Bulgarian Association for Intelligent Systems and the Bulgarian Society of Robotics. ### Academic collaboration Scientific and educational cooperation with the Faculty of Biology at Sofia University. ### Work planned for the next 24 months - A locked FOLKLORE Clinical v1 configuration with defined intended purpose and module boundaries. - Independent evaluation at 3-5 sites in at least two EU/EEA countries. - Two production API/LIS integrations and entry into 3-5 priority EU/EEA markets. - 20-30 paying or contracted accounts and EUR 0.5-0.8m annualised revenue. - A configuration ready to enter IVDR conformity assessment. ## The current round funds validation and European deployment We have financed the product, European infrastructure, laboratory deployment and retrospective evidence programme without institutional investment. We are seeking financing to expand our commercial reach across Europe and support additional laboratory deployments. - Completed foundationFounder-financed product and evidence programmeOperational RUO platform, laboratory workflow, four evidence cohorts and initial commercial agreements. Completed foundation ### Founder-financed product and evidence programme Operational RUO platform, laboratory workflow, four evidence cohorts and initial commercial agreements. - Current financingValidation and European deploymentIndependent validation, regulated product preparation, operational hiring and European deployment. Current financing ### Validation and European deployment Independent validation, regulated product preparation, operational hiring and European deployment. - Months 18-24Growth financingIntended for broader European growth after independent evidence, initial recurring revenue and progress towards conformity assessment. Months 18-24 ### Growth financing Intended for broader European growth after independent evidence, initial recurring revenue and progress towards conformity assessment. ## Current team and planned hires We currently operate with founder leadership, scientific governance and external validation support. With financing, we plan to add clinical, regulatory, engineering and commercial capacity. Leadership ### Vladimir Mitev Founder & CEO Vladimir leads product architecture, software engineering, infrastructure, company strategy and commercial deployment. Scientific governance ### Academician Prof. Draga Toncheva, MD, DSc Founding Chair, Scientific & Clinical Advisory Board Draga Toncheva provides scientific direction, clinical validation oversight and human governance across genetics, genomics and molecular biology. Validation contributors ### External evidence review Molecular geneticists review classifier behaviour against case evidence, identify discrepancies and help refine the evidence workflow. Daniela Maneva and Spasimir Shishinyov Planned hires ### Capacity to be added with financing The next hires will add the capabilities needed for broader European deployment. #### Platform / DevOps Build, scale and secure our production platform and data infrastructure. #### Clinical genetics Strengthen clinical interpretation, variant curation and evidence standards. #### Quality / Regulatory Establish and maintain quality systems and regulatory compliance. #### Commercial Europe Develop partnerships, laboratory onboarding and market access across Europe. ## Helena Bioinformatics Our legal and operating details are listed below for investors, financing programmes and external company databases. Legal name Helena Bioinformatics EOOD Headquarters Sofia, Bulgaria Stage Seed Product FOLKLORE Sector Bioinformatics, clinical genomics, health software Business model B2B software, usage-based and volume access Geographic focus Europe, with the Balkans as the initial commercial region Regulatory status Research Use Only; not CE-marked EU Participant Identification Code 866588592 Find us on ### Links and cited sources - [View pitch deck](https://folklore.helena.bio/pitch-deck) - [Explore FOLKLORE](https://folklore.helena.bio/) - [Cohort 1Pathogenic Variant Concordance Study20/20 cases passed every primary criterionFolklore detected every target finding, assigned the correct gene and achieved 100% FULL plus CLINICAL concordance.](https://www.helena.bio/reports/validation/folklore-cohort-1-public-report.pdf) - [Study details and methodology](https://folklore.helena.bio/studies/cohort-1) - [Public report on OSF](https://osf.io/kvj8e/) - [DOI 10.17605/OSF.IO/KVJ8E](https://doi.org/10.17605/OSF.IO/KVJ8E) - [Cohort 2Multi-Source Performance Characterization19/20 target variants detectedFolklore assigned every evaluable target to the correct gene. No opposite-direction classification occurred. The one missed target was absent from the supplied variant input.](https://www.helena.bio/reports/validation/folklore-cohort-2-public-report.pdf) - [Study details and methodology](https://folklore.helena.bio/studies/cohort-2) - [Public report on OSF](https://osf.io/rtbk8/) - [DOI 10.17605/OSF.IO/RTBK8](https://doi.org/10.17605/OSF.IO/RTBK8) - [Cohort 3Real-World Performance Study98.9% in-scope variant detectionFolklore detected 184 of 186 shared variants. Independent forensic review found no classifier-logic defect, and all 49 AI-bearing reports passed the gene-disease name faithfulness check.](https://www.helena.bio/reports/validation/folklore-cohort-3-public-report.pdf) - [Study details and methodology](https://folklore.helena.bio/studies/cohort-3) - [Public report on OSF](https://osf.io/k8uwf/) - [DOI 10.17605/OSF.IO/K8UWF](https://doi.org/10.17605/OSF.IO/K8UWF) - [Cohort 4Deterministic Classification Performance Study67.0% FULL + CLINICAL agreementAll 100 cases completed under one locked classifier version. Forensic review of 19 priority variants across 16 cases separated a bounded correction programme from reference-data gaps, manual-evidence limits and defensible inter-laboratory differences.](https://www.helena.bio/reports/validation/folklore-cohort-4-public-report.pdf) - [Study details and methodology](https://folklore.helena.bio/studies/cohort-4) - [Public report on OSF](https://osf.io/nphw8/) - [DOI 10.17605/OSF.IO/NPHW8](https://doi.org/10.17605/OSF.IO/NPHW8) - [Meet the team and advisory board](https://www.helena.bio/team) - [Helena Bioinformatics on Crunchbase](https://www.crunchbase.com/organization/helena-bioinformatics) - [Helena Bioinformatics on Dealroom](https://app.dealroom.co/companies/helena_bioinformatics) - [Helena Bioinformatics on Vestbee](https://app.vestbee.com/s/22afe04e39f0450ea8c827775d33fbc0) - [Helena Bioinformatics on F6S](https://www.f6s.com/helena-bioinformatics) - [Helena Bioinformatics on Gust](https://gust.com/companies/helena-bioinformatics) - [Helena Bioinformatics on Tracxn](https://platform.tracxn.com/a/d/company/590515c6e4b044ce9d7e6045/helenabioinformatics) - [Helena Bioinformatics on EuroQuity](https://www.euroquity.com/en/company/helena-bioinformatics) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Partners | Helena Bioinformatics Source: https://www.helena.bio/partners Canonical: https://www.helena.bio/partners Description: Helena Bioinformatics collaborates with laboratories, distributors, universities, and research organisations in clinical genomics. # Partnerships built for clinical genomics We collaborate with laboratories, distributors, universities and research organisations to deliver trusted genomic interpretations that scale. Together, we build the scientific, clinical and commercial foundations for genomic interpretation at scale. ## CellGenetics Sofia / Plovdiv / Varna Whole-genome sequencing analysis and interpretation for clinical diagnostics. CellGenetics is our founding clinical partner and the first laboratory to implement Folklore in its routine workflow. Their expertise in clinical genomics, combined with deep experience in whole-genome sequencing analysis and interpretation, played a key role in shaping and validating the platform. ## ELTA 90 Bulgaria Laboratory technology and diagnostics distributor with expertise in NGS and molecular genetics. ELTA 90 is Helena Bioinformatics' exclusive distribution partner in Bulgaria. With extensive experience in laboratory technologies, next-generation sequencing and molecular genetics, ELTA 90 supports laboratories with products, training, implementation and technical service. ## Faculty of Biology, Sofia University Sofia, Bulgaria Scientific and educational partnership in genomics, bioinformatics and biomedical data analysis. The Faculty of Biology at Sofia University "St. Kliment Ohridski" is Helena Bioinformatics' scientific and educational partner in genomics, bioinformatics and biomedical data analysis. The collaboration brings together the Faculty's academic and research expertise with Helena's computational infrastructure and work in clinical genomics. The agreed scope includes genomic and metagenomic analysis, molecular biology, artificial intelligence, biological data standards, reproducibility and the evaluation of digital methods used in life sciences and medicine. ## Institute of Information and Communication Technologies, Bulgarian Academy of Sciences Sofia, Bulgaria Research in intelligent systems, AI methods and data-driven decision support. The Institute of Information and Communication Technologies at the Bulgarian Academy of Sciences is Helena Bioinformatics' research collaboration partner in intelligent systems and applied AI. Our collaboration focuses on methods, data structures and computational approaches relevant to clinical interpretation, decision support and biomedical informatics. ## Bulgarian Association of Intelligent Systems Sofia, Bulgaria Professional association advancing intelligent systems and artificial intelligence across research and society. The Bulgarian Association on Intelligent Systems (BAIS) is a non-profit professional association supporting the development and application of intelligent systems and artificial intelligence. Its work connects research, education, industry, healthcare and public institutions through studies, expert analysis, conferences, training and national and international initiatives. ## Bulgarian Society of Robotics Sofia, Bulgaria Research in robotics, automation and intelligent medical systems. The Bulgarian Society of Robotics (BSR) conducts research and applied activities in robotics, automation and intelligent systems, including emerging applications in medicine. Our collaboration explores how genomic and clinical information can support automated, data-driven and intelligent medical technologies. ## Starting a partnership discussion A partnership discussion starts with the work the organisations want to do together. Laboratories can describe an interpretation workflow; research teams can identify a scientific question and public evidence; distributors can explain the organisations and region they serve. The contact route above lets you introduce that context and the people involved. Our partner profiles explain the specific relationship rather than treating every collaboration as the same kind of endorsement. CellGenetics is presented as our founding clinical partner. Its profile provides further information about the laboratory and the collaboration. For the evidence behind Folklore, consult the public studies and their stated evaluation boundaries alongside the product documentation. When contacting us, include your organisation, a relevant public link and the next conversation you would find useful. Initial discussions can use a description of the workflow without sharing patient records or confidential research data. ### Links and cited sources - [Full profile](https://www.helena.bio/partners/cellgenetics) - [Visit website](https://cellgenetics.bg/) - [Visit website](http://www.elta90.com/?lang=en) - [Visit website](https://www.uni-sofia.bg/index.php/eng/the_university/faculties/faculty_of_biology2) - [Visit website](https://iict.bas.bg/en/index.html) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: CellGenetics - Founding Clinical Partner | Helena Bioinformatics Source: https://www.helena.bio/partners/cellgenetics Canonical: https://www.helena.bio/partners/cellgenetics Description: CellGenetics is Helena Bioinformatics's founding clinical partner, with work in whole-genome sequencing interpretation, rare disease, oncology, cardiology, neurology, and reproductive health. # CellGenetics Founding Clinical Partner ## About CellGenetics CellGenetics specializes in the analysis and interpretation of whole-genome sequencing data. Its diagnostic work includes rare disease, oncology, cardiology, neurology, reproductive health, and pharmacogenomics. CellGenetics combines bioinformatic analysis with interpretation by board-certified clinicians. The laboratory operates from three locations in Bulgaria for sample collection and processing. ## Clinical Domains CellGenetics covers seventeen clinical domains in genetic diagnostics through whole-genome sequencing and targeted panels. ## Team The CellGenetics team includes physicians board-certified in medical genetics, with extensive experience in counseling families affected by rare diseases, and molecular geneticists specialized in bioinformatic analysis of whole-genome sequencing data. The laboratory has contributed to peer-reviewed international publications, including the identification of novel genetic findings in Bulgarian patients. ## Nationwide Coverage CellGenetics operates from three locations - Sofia, Plovdiv, and Varna - with a unified national infrastructure for sample collection and processing. ### Sofia Headquarters 15 Tsar Petar Street ### Plovdiv Regional office 5 Sredets Street, floor 2, office 1 ### Varna Regional office 3 Dospat Street ## Standards & Methodology Helena evaluation uses the standards listed below. CellGenetics clinical reports are signed by the laboratory's medical geneticists and molecular geneticists and serve as the comparator for the current cohorts. ISO 15189:2022 ### Medical Laboratory Quality Management International standard for quality and competence of medical laboratories. Defines requirements for quality management systems, examination procedures, and result reporting in clinical diagnostic settings. EU IVDR (2017/746) ### In Vitro Diagnostic Regulation European Union regulation governing the safety, performance, and clinical evidence requirements for in vitro diagnostic medical devices. Applies to laboratory-developed tests and bioinformatics pipelines used in clinical diagnostics. AMP/CAP (2018) ### Bioinformatics Pipeline Validation Joint Association for Molecular Pathology and College of American Pathologists guidelines for validating clinical next-generation sequencing bioinformatics pipelines, covering analytical validation, performance characteristics, and continuous quality monitoring. ACMG/AMP (2015) ### Sequence Variant Interpretation Standards and Guidelines for the Interpretation of Sequence Variants from the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. The framework Helena automates with the Bayesian point system (Tavtigian et al. 2018) and ClinGen SVI calibrated thresholds (Pejaver et al. 2022). ## Clinical Validation ### Validation in progress Validation cohorts are running with CellGenetics as the reference standard. Cases are selected from the laboratory's patient registry to maximize coverage across clinical domains, variant types, zygosity categories, and complexity levels. All concordance criteria and acceptance thresholds are defined before processing begins. Detailed validation results will be published here upon completion and clinical partner approval. ## Contact ### CellGenetics 0700 17 578 info@cellgenetics-lab.com cellgenetics.bg ### Helena partnership inquiries For partnership inquiries with Helena Bioinformatics, please contact us directly. ### Links and cited sources - [info@cellgenetics-lab.com](mailto:info@cellgenetics-lab.com) - [cellgenetics.bg](https://cellgenetics.bg/) - [Contact Helena](https://www.helena.bio/contact) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Privacy Notice | Helena Bioinformatics Source: https://www.helena.bio/privacy Canonical: https://www.helena.bio/privacy Description: How Helena processes personal data as controller and processor. # Privacy Notice Version 1.0 · Effective: 14 August 2026 This Privacy Notice explains how Helena Bioinformatics EOOD processes personal data when acting as a controller for its websites, accounts, communications and business operations. It also explains the separate situation in which Helena processes clinical or genomic data on documented instructions from a customer that acts as controller. This Notice is information under Articles 13 and 14 GDPR. Acknowledging it does not create consent and does not replace a Data Processing Agreement or a specific consent where consent is the applicable legal basis. ## 1. Controller and privacy contact For controller activities, the controller is Helena Bioinformatics EOOD, UIC 208673837, 14 Tsar Ivan Asen II Street, floor 1, apartment 1, 1142 Sofia, Bulgaria. Contact: privacy@helena.bio. The privacy email is Helena’s data-protection contact. It should not be interpreted as a formal designation of a data protection officer unless Helena separately publishes such a designation. ## 2. Our roles Helena acts as controller for account registration, identity and access management, legal acceptance evidence, security records, support, demo and contact requests, website operation, product communications, billing administration and its own legal compliance. A laboratory, healthcare provider, researcher or other customer normally acts as controller for patient, sample, phenotype, genomic and case data submitted to a private workspace. Helena acts as processor for that data under the customer’s documented instructions and the applicable DPA. Data subjects should normally direct requests about such data to the relevant customer. ## 3. Data we process as controller Identity and account data: name, professional and organizational affiliation, role, email, language, account profile, organization relationships, status and password hash. Helena does not store the account password in readable form. Contract and compliance data: accepted legal-document version and digest, acceptance time and source, contracts, approvals, communications and records needed to establish or defend legal claims. Security and technical data: IP address where logged by infrastructure, device and browser information, session and token identifiers, timestamps, request identifiers, authentication events, audit events, error and diagnostic data. Communication and business data: contact, demo, support and partnership requests; correspondence; scheduling information; billing contact, customer and subscription identifiers when paid billing is enabled. Payment-card details are collected by Stripe and are not stored by Helena. Public-service interaction data: public search requests and bounded questions sent to the public AI assistant. Users must not submit patient identifiers or confidential case information to public features. Optional integration data: when a user deliberately links a supported external account, Helena may process the identifier and encrypted connection state needed for that integration. The product will identify the integration and may require additional notice before activation. ## 4. Customer-controlled clinical and genomic data Depending on customer configuration, processor data may include pseudonymized sample identifiers, VCF and other variant data, phenotype terms, family relationships, analysis parameters, clinical notes supplied by the customer, generated evidence and reports, provenance and access records. Genetic and health data are special-category data. The customer, not Helena, determines the applicable Article 6 and Article 9 GDPR bases and must ensure that direct identifiers are not uploaded unless expressly agreed. Helena processes this data only under documented instructions, except where Union or Member State law requires otherwise. ## 5. Purposes and legal bases for controller processing Contract and pre-contract steps (Article 6(1)(b)): create and operate accounts, provide requested services, authenticate users, handle support and administer subscriptions. Legal obligations (Article 6(1)(c)): tax and accounting records, legally required security or breach actions, regulatory cooperation and responses to lawful requests. Legitimate interests (Article 6(1)(f)): secure the Service, prevent abuse, maintain auditability, diagnose faults, improve reliability, respond to business contacts and establish or defend legal claims. Helena balances these interests against individual rights and applies minimization and access controls. Consent (Article 6(1)(a)): non-essential analytics or marketing communications where consent is required. Consent is optional, may be withdrawn at any time and does not affect earlier lawful processing. Essential authentication and security processing is not based on consent. ## 6. Sources and required information We receive data directly from you, from an organization that invites or administers you, from a distributor or laboratory establishing an approved relationship, from service providers operating on our behalf, and automatically from devices and systems when the Service is used. Information marked as required is necessary to create and secure an account or perform a contract. Without it, Helena may be unable to provide the requested account or feature. Optional analytics and marketing choices do not affect essential access. ## 7. Recipients and service providers Hetzner Online GmbH provides EU server and infrastructure hosting. Vercel Inc. provides delivery and hosting for public websites and frontend applications and may receive network and usage metadata, but is not intended to receive private genomic case content. Resend provides transactional email delivery and receives recipient email, message content and delivery metadata. Google services may process consent-based website analytics and information needed to handle contact requests or schedule demonstrations. OpenAI processes bounded public assistant content; the public assistant is not intended for patient or private case data and provider storage is disabled in Helena’s configured request. Stripe will process billing identity, transaction and payment information as an independent controller and/or processor under its own terms when paid billing is enabled. Payment-card data is entered into Stripe-hosted surfaces. MEGA or another user-selected integration processes data only when the user enables that integration. Professional advisers, auditors, insurers, public authorities and courts may receive data where necessary and lawful. We do not sell personal data. A current provider register should be reviewed together with this Notice because providers and legal entities may change. ## 8. International transfers Private genomic workloads are designed to run on Helena-controlled infrastructure in the European Economic Area. Public website, email, analytics, support, payment and public-AI providers may process limited personal data outside the EEA. Where a recipient is outside the EEA and no adequacy decision applies, Helena uses an applicable transfer mechanism such as the European Commission’s Standard Contractual Clauses together with supplementary measures where required. You may request information about the relevant safeguards at privacy@helena.bio. No customer-controlled clinical data is sent to an external AI provider unless the customer agreement, product mode, DPA and required safeguards expressly allow it. ## 9. Retention Account and organization data is kept while the account is active and afterwards only for the period necessary for closure, security, legal, tax and claims purposes. Verification and abandoned pending-account data is removed or anonymized according to an operational retention schedule. Legal acceptance and contract evidence is retained for the applicable limitation period and any additional period required to establish compliance, subject to access restrictions and applicable erasure rights. Security, audit and diagnostic records are retained according to risk-based schedules and are not represented as having a fixed period unless that schedule is operationally enforced. Customer-controlled clinical and genomic data is retained for the period selected in the applicable order or DPA. Deletion from active systems and backup expiry follow documented technical procedures. Helena will not promise automatic permanent deletion within a particular period unless that period is configured and supported for the relevant service. Provider records may remain under the provider’s lawful retention rules. Helena periodically reviews retention and documents any legal hold. ## 10. Security Helena applies risk-appropriate technical and organizational measures, including access control, least privilege, credential hashing, encryption in transit, network restrictions, logging, monitoring, backup and recovery controls, vulnerability management and incident procedures. The precise measures applicable to customer-controlled data are stated in the DPA and may evolve without reducing the overall level of protection. No system is completely secure. Users must protect their credentials and promptly report suspected compromise. Helena will notify controllers and individuals of personal-data breaches where and within the periods required by law and contract. ## 11. Cookies and similar technologies Authenticated services use strictly necessary storage for security, authentication, language and session continuity. Public websites may use consent-based analytics. Non-essential analytics is disabled until the visitor chooses to allow it, and the choice can be changed through the analytics settings. We do not intentionally send genomic search expressions, variant coordinates, patient data, authenticated case content or result URL parameters to website analytics. ## 12. Automated processing Helena uses automated analysis and prioritization to support qualified users. Helena does not use controller-side account data to make a solely automated decision that produces legal or similarly significant effects on an individual. Clinical users remain responsible for review and decisions based on analysis outputs. If this changes, Helena will provide the information and safeguards required by Articles 13, 14 and 22 GDPR before the processing begins. ## 13. Your rights Subject to applicable conditions, you may request access, correction, erasure, restriction, portability or objection, and may withdraw consent where processing is based on consent. You may also complain to the Bulgarian Commission for Personal Data Protection, 2 Prof. Tsvetan Lazarov Blvd., 1592 Sofia, Bulgaria, or another competent supervisory authority. Send controller-side requests to privacy@helena.bio. Helena may verify identity and normally responds within one month, subject to lawful extensions. For patient or customer-workspace data, contact the relevant laboratory or customer as controller; Helena will assist it under the DPA. ## 14. Children The account Service is not directed to children and account holders must be at least 18. Patient data concerning children may be processed only when submitted by an authorized controller with an appropriate legal basis and safeguards. ## 15. Changes and versioning Each version of this Notice is identified by version, language and effective date and is preserved in an immutable archive. Material changes are communicated to registered users. Where acknowledgement is required, Helena records the exact document version and digest presented to the user. A new Notice does not retroactively alter the lawfulness of earlier processing. If a new purpose requires consent or another action, Helena will obtain it separately before that processing. ## 16. Contact Questions, requests and requests for transfer-safeguard information may be sent to privacy@helena.bio or by post to the registered address above. ### Links and cited sources - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Terms of Use | Helena Bioinformatics Source: https://www.helena.bio/terms Canonical: https://www.helena.bio/terms Description: Terms governing access to and use of Helena and Folklore. # Terms of Use Version 1.0 · Effective: 14 August 2026 These Terms of Use (the “Terms”) are a binding agreement between Helena Bioinformatics EOOD (“Helena”, “we”, “us”) and the person or legal entity that accesses or uses Helena, Folklore, their applications, APIs, documentation and related services (together, the “Service”). By creating an account, accepting these Terms or using the Service, you confirm that you have read and agree to these Terms. If you act for an organization, you confirm that you have authority to bind it. If you do not agree, do not create an account or use the Service. ## 1. Provider and contact details The Service is provided by Helena Bioinformatics EOOD, UIC 208673837, with registered office at 14 Tsar Ivan Asen II Street, floor 1, apartment 1, 1142 Sofia, Bulgaria. Legal questions may be sent to legal@helena.bio and security incidents to security@helena.bio. ## 2. Eligibility and account profiles The Service is intended for qualified genetics, molecular biology, bioinformatics, laboratory and healthcare professionals, researchers acting within an authorized project, and organizations employing or engaging such persons. It is not offered directly to patients for diagnosis or treatment decisions. An Individual account is for a qualified professional working independently and creates a private workspace. A Laboratory account creates a shared organizational workspace and requires an authorized administrator. A Distributor account is only an application for a commercial partner relationship and remains restricted until Helena approves a separate distributor agreement. You must be at least 18 years old and legally capable of entering into these Terms. You must provide accurate information and keep it current. Helena may request evidence of professional qualification, authority or organizational identity before enabling clinical or commercial capabilities. ## 3. Account security and administration You must keep credentials confidential, use the account only for its assigned person, maintain accurate administrator and contact information, and notify security@helena.bio promptly of suspected compromise or unauthorized access. Organization administrators control membership and roles within their organization. They are responsible for inviting only authorized users, reviewing access regularly and removing access promptly when it is no longer needed. Shared credentials are prohibited. ## 4. The Service and plan availability Helena provides software-assisted genomic and biomedical analysis functions, which may include variant annotation and classification support, phenotype matching, literature retrieval, prioritization, reporting and collaboration tools. Features may differ by plan, account profile, configuration, geography and release stage. Version 1.0 supports a free plan without a payment card. Paid plans, usage charges, taxes, renewal, cancellation and refund terms do not apply unless you separately select a paid offer whose ordering and billing terms are presented before purchase. Helena may introduce, change or discontinue plans prospectively, but will not impose a charge without an affirmative purchase step. ## 5. Professional and clinical responsibility The Service is a professional decision-support and research tool. Outputs may be incomplete, uncertain, outdated or incorrect and must be reviewed against the underlying evidence, applicable professional standards, laboratory procedures and the complete clinical context by a suitably qualified person. The Service does not itself establish a diagnosis, prescribe treatment, determine patient management or replace professional judgment. You remain responsible for validation, confirmation, reporting and all clinical or research decisions. Clinically significant findings should be independently verified where required by applicable standards. The regulatory status of particular features depends on their intended purpose, configuration and use. Nothing in these Terms is a representation that every feature is, or is not, a medical device. Helena will identify any separately regulated feature and its applicable instructions when required. ## 6. Customer data and lawful authority You retain your rights in data, files and materials submitted to the Service (“Customer Data”). You grant Helena the limited rights necessary to host, copy, transmit, analyze and otherwise process Customer Data solely to provide, secure, support and maintain the Service and to comply with law. You must have a valid legal basis and all necessary notices, permissions, approvals and documented instructions before submitting personal, genetic, health or other confidential data. Patient data must be pseudonymized before upload unless Helena has expressly agreed otherwise in writing. Do not upload direct identifiers or data outside the agreed scope. Where an organization determines the purposes and means of processing personal data and Helena processes it on that organization’s behalf, the Data Processing Agreement forms part of the applicable service agreement. Individual account holders who act as controllers must contact Helena before processing patient data unless an applicable DPA has been executed. ## 7. Organization and distributor boundaries Users within the same organization may access shared cases and resources according to assigned roles. The organization is responsible for its users and their use of the Service. A distributor relationship does not give the distributor access to a laboratory’s users, cases, patient data or clinical workspace. Commercial ownership, invoicing and organization relationships are separate from membership and clinical authorization. Any distributor rights and charges are governed by a separate written distributor agreement. ## 8. Acceptable use You must not use the Service unlawfully; submit data without authority; identify or attempt to re-identify a person from anonymized or pseudonymized data; use outputs as an unreviewed diagnosis or treatment instruction; infringe intellectual property or confidentiality rights; introduce malware; probe or bypass security controls; disrupt the Service; scrape at unreasonable volume; misrepresent output provenance; or enable another person to do so. You must not reverse engineer or attempt to extract source code, models, non-public datasets or security mechanisms except to the limited extent such restriction is prohibited by mandatory law. Reasonable, good-faith security research must be reported to security@helena.bio and must avoid patient data, service disruption and unauthorized disclosure. ## 9. Artificial intelligence features Some features may generate or summarize content using statistical or artificial-intelligence models. Such output is probabilistic, may resemble output produced for others and is not guaranteed to be unique, complete or factually correct. Do not place direct patient identifiers or confidential information into a public or external-provider AI feature. Helena will identify protected and public AI boundaries in the product where relevant. You must review citations and source evidence before relying on generated content. ## 10. Intellectual property and feedback Helena and its licensors retain all rights in the Service, software, interfaces, documentation, methods, branding and underlying technology. Subject to these Terms, Helena grants you a limited, non-exclusive, non-transferable and revocable right to use the Service during the applicable account term for its intended purpose. You may use and include analysis results in your authorized professional or research workflow, subject to third-party database licences, attribution requirements and patient rights. You grant Helena a perpetual right to use feedback you voluntarily provide without identifying you or disclosing Customer Data. ## 11. Third-party services and data sources The Service may interoperate with third-party services and draw on public or licensed scientific sources. Those sources may change, contain errors or be subject to separate terms. Helena is not responsible for a third party’s independent service, but remains responsible for its own obligations and for processors it appoints as required by law. ## 12. Confidentiality Each party must protect the other party’s non-public business, technical and clinical information using at least reasonable care and use it only for the relationship. This obligation does not apply to information demonstrably public without breach, already lawfully known, independently developed or lawfully received without restriction. Required disclosure is permitted after advance notice where legally allowed. ## 13. Availability, changes and support Helena aims to operate the Service reliably but does not guarantee uninterrupted or error-free availability. Maintenance, security events, dependency failures and force-majeure events may affect access. Support and any service level commitment apply only if stated in a separate written order or agreement. Helena may update the Service for security, compliance, interoperability or product development. We will give reasonable notice when a material change materially reduces an active paid service, unless urgent action is needed to address security, law or third-party dependency risk. ## 14. Suspension and termination You may stop using the free Service and request account closure at any time. Either party may terminate an applicable service agreement as provided in that agreement. Helena may suspend access immediately where reasonably necessary to address a security threat, unlawful use, material breach, non-payment under a paid order or risk to patients, users or the Service. Where practicable, Helena will give notice and an opportunity to remedy. On termination, access ends and Customer Data is returned or deleted according to the applicable DPA, retention policy and legal obligations. Provisions intended by their nature to survive, including confidentiality, intellectual property, accrued payment obligations, liability and dispute provisions, continue. ## 15. Warranties and disclaimers Each party warrants that it has authority to enter into these Terms. Helena will provide the Service with reasonable skill and care. Except for express commitments in these Terms or a signed agreement, and to the extent permitted by law, the Service is provided “as is” and Helena disclaims implied warranties of merchantability, fitness for a particular purpose, non-infringement and that every result will be accurate or suitable for clinical use. Mandatory rights are not excluded. ## 16. Liability Nothing excludes or limits liability that cannot lawfully be excluded, including liability for fraud, wilful misconduct, death or personal injury caused by negligence where applicable, or obligations under data-protection law that cannot be limited by contract. Subject to the preceding sentence, neither party is liable for indirect or consequential loss, loss of profit, revenue, goodwill or anticipated savings. Helena’s aggregate contractual liability in a twelve-month period is limited to the greater of EUR 500 and the fees paid or payable for the Service in that period. This limit does not apply where a separate signed agreement establishes a different limit. The limitations apply only to the maximum extent permitted by applicable law and do not restrict mandatory consumer rights. The Service is not currently offered as a consumer medical service. ## 17. Changes to these Terms Helena may issue a new version of these Terms. Each version is identified by version, language and effective date. We will preserve prior versions and notify registered users of material changes. Where acceptance is required, the new version will not bind continued use until the applicable effective date and acceptance workflow, except where law requires immediate action. A prior acceptance does not count as acceptance of a later version. If you do not accept a required new version, access to affected features may be restricted and you may close the account subject to data-return and retention rules. ## 18. Governing law and general terms These Terms are governed by Bulgarian law and directly applicable European Union law, without regard to conflict-of-law rules. The competent courts in Sofia, Bulgaria have jurisdiction, except where mandatory law grants another forum. If a provision is unenforceable, it will be limited to the minimum extent necessary and the remaining provisions remain effective. Failure to enforce a provision is not a waiver. You may not assign these Terms without Helena’s written consent; Helena may assign them as part of a reorganization or transfer of the Service while preserving applicable rights. The English and Bulgarian versions are intended to correspond; unless mandatory law requires otherwise, the English version prevails in case of inconsistency. ### Links and cited sources - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Google API Data Policy | Helena Bioinformatics Source: https://www.helena.bio/google-api-data-policy Canonical: https://www.helena.bio/google-api-data-policy Description: How Helena Bioinformatics accesses, uses, stores, shares, and protects Google Drive data in Folklore. Google API Services # Google API Data Policy Last updated: 27 August 2026 This policy explains how Helena Bioinformatics EOOD uses Google Drive data when a Folklore user explicitly connects a Google account. ## Access requested Folklore requests the read-only Google Drive scope https://www.googleapis.com/auth/drive.readonly. It lets the user browse file metadata in My Drive and explicitly approved Shared Drives and download a selected binary genomic file for import. The narrower drive.file scope is not sufficient because Folklore must read genomic files that already exist in user-selected Drive roots and were not created by Folklore or individually opened through the Google Picker. Folklore does not create, edit, rename, move, share, or delete Google Drive content. ## Use, storage, and deletion Drive access is initiated by the user. Folklore lists only the configured roots and imports only a file selected by the user. Google Docs, shortcuts, trashed objects, link-only objects, and files that cannot be downloaded are excluded. OAuth tokens are encrypted at rest and used through short-lived operation leases. Disconnecting Google Drive removes the stored connector credentials. A genomic file that the user imports becomes customer-controlled case data and is handled under the Folklore Privacy Notice, contractual retention settings, and the customer's instructions. ## Sharing and AI/ML Raw files downloaded from Google Drive are not sent to an external AI provider. If a user separately invokes Folklore's private case assistant, a bounded derived case context may be sent directly to the OpenAI API to answer that user's request. The private case-assistant path uses OpenAI's standard API in a dedicated project; it does not use an aggregator or model hub. It excludes raw files, full genomes, direct identifiers, demographics, and unrestricted clinical notes. Helena Bioinformatics does not use raw or derived Google Workspace API data to develop, improve, or train generalized AI or machine-learning models, and does not sell it or use it for advertising. The use of information received from Google APIs adheres to the Google API Services User Data Policy, including the Limited Use requirements. ## Contact and related policy Questions can be sent to privacy@helena.bio. See the Privacy Notice for the broader privacy framework. ### Links and cited sources - [privacy@helena.bio](mailto:privacy@helena.bio) - [Privacy Notice](https://www.helena.bio/privacy) - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/) --- ## Page: Ask Helena Bioinformatics Source: https://www.helena.bio/ask Canonical: https://www.helena.bio/ask Description: Get source-cited answers about Helena Bioinformatics, its products, research and public documentation. Helena Bioinformatics # Ask Helena Ask Helena Bioinformatics about the company, Folklore, Evidence, or another product represented in its approved public sources. Every result points back to the public pages that support it. Ask Helena does not provide medical advice or live variant interpretation. ## Using Ask Helena Ask Helena answers questions about Helena Bioinformatics and its products from approved public sources. You can ask what a product does, where its documentation is published, how to find a public study or what a published method covers. The service provides source citations and identifies limits when the available material does not support a complete answer. A focused question helps keep the answer tied to the right product and source. For example, name Folklore when asking about its variant interpretation workflow, or name a research paper when asking about its methods. Questions about Noodle can concern biomedical literature discovery; questions about Evidence can concern public study records. Ask one question at a time and use the cited documents to inspect the underlying detail. Follow citations before relying on a claim. A research preprint, a methodological proposal and a completed evaluation have different purposes. Check the source date, version, study population and stated limitations. If the response provides sources without a supported answer, read those sources or reformulate the question around a specific public document. An unavailable answer does not establish that a product has or lacks a capability. This is a public company-information service. Do not submit patient information, clinical cases, variant or phenotype data, family information, VCF content or medical reports. Use the public documentation to understand product scope and contact Helena for questions that require a discussion with the team. Ask Helena does not replace specialist interpretation or a clinical consultation. ### Links and cited sources - [https://www.helena.bio/](https://www.helena.bio/) - [Ask Helena](https://www.helena.bio/ask) - [Blog](https://blog.helena.bio/) - [Publications](https://www.helena.bio/research) - [Studies](https://folklore.helena.bio/studies) - [Research and Methods](https://www.helena.bio/#work) - [Evidence Admissibility](https://zenodo.org/records/21346784) - [RWP Framework](https://doi.org/10.3389/fgene.2026.1925492) - [Team](https://www.helena.bio/team) - [Investors](https://www.helena.bio/investors) - [Partners](https://www.helena.bio/partners) - [Careers](https://www.helena.bio/careers) - [Privacy Policy](https://www.helena.bio/privacy) - [Google API Data Policy](https://www.helena.bio/google-api-data-policy) - [Terms of Service](https://www.helena.bio/terms) - [DPA](https://www.helena.bio/dpa) - [DPIA](https://www.helena.bio/dpia) - [LinkedIn](https://www.linkedin.com/company/helena-bioinformatics/) - [YouTube](https://www.youtube.com/@helenabioinformatics) - [Bluesky](https://bsky.app/profile/helena.bio) - [Genomic Social](https://genomic.social/@helena_bio) - [Medium](https://medium.com/helena-bioinformatics) - [GitHub](https://github.com/orgs/helena-bioinformatics/)