Reproducible records
Projects record the data source, preprocessing, split design, model and software versions, comparison methods, assumptions, and limitations needed to assess the result.
Public records / research
Research practice
VHL conducts research toward computational models of human biology across medicine, transcriptomics, bioinformatics, computational biology, and machine learning. Current public work is focused on RNA methods and evaluation. Its standard is to keep sufficient records for others to understand the data, methods, assumptions, citations, conflicts, limitations, and material corrections behind a claim.
Clinical scope
VHL is not a medical provider and does not provide diagnosis, treatment, prescriptions, or patient-specific medical advice.
01 / Standards
What readers should expect from VHL work.
These standards apply whether a result is public, under review, or shared privately with a researcher.
Contribution and responsibility
Authorship follows actual scholarly contribution and responsibility. Membership, seniority, funding, or advisory status alone does not create authorship.
Human accountability
Material use of generative AI or automated research systems is disclosed when required. Human researchers remain responsible for accuracy, integrity, privacy, and compliance.
Leakage stated and controlled
Biological-model results identify the split unit and account for repeated donors, cohorts, studies, sites, and batches that could inflate apparent generalization.
Material changes are recorded
Errors or new evidence that change a result, method, or interpretation should produce a dated correction or revision.
Identity, not sponsorship
Professional affiliations identify individuals. They do not make an employer or institution responsible for VHL work unless a written agreement says so.
02 / Human data
Permission, purpose, and review.
Public availability is not a shortcut around repository terms, ethics review, data-use agreements, privacy requirements, or a member’s institutional duties.
Human-subject, biospecimen-derived, controlled-access, or individual-level clinical data require the approvals, permissions, and safeguards that apply to that dataset and research use.
VHL does not use patient records, non-public military medical data, internal institutional data, company-confidential information, accounts, equipment, or facilities without prior authorization.
03 / Rights and release
Participation alone does not transfer intellectual property.
Project rights depend on actual contribution, law, institutional policies, funding conditions, and written project terms.
Existing work stays with its owner
Joining VHL does not by itself transfer a person’s or institution’s background intellectual property to the group.
Rights are checked before release
Code, models, datasets, and documents receive an open license only after ownership and third-party restrictions have been reviewed.
Major rights need collective approval
Patents, exclusive licenses, assignments, commercial priority rights, and material data or institutional agreements require the approval specified in the bylaws.