The initial beta is focused on AncestryDNA-style SNP-array raw data.
About DNA_CHECKS
DNA_CHECKS
An independently developed personal genetics project for interpreting existing raw DNA data using curated scientific evidence and reproducible analysis pipelines.
The project focuses on transparent evidence, cautious interpretation and keeping technical limitations visible.
Independent project
Small by design
DNA_CHECKS is currently independently developed rather than operated by a large genetics company or health-tech platform.
The aim of the beta is not to build a large collection of people's DNA. It is to test whether a transparent, evidence-linked personal genetics report is useful enough that people want to use it, improve it and support continued development.
Feedback, scientific criticism, bug reports, feature ideas and technical contributions are welcome.
Current boundaries
Beta means there are boundaries
The first public version is intentionally narrow. Features should expand when the scientific evidence and engineering justify them.
Consumer arrays measure selected positions and cannot support every possible analysis.
Important clinical findings may require confirmation using appropriate clinical-grade testing.
A report reflects the evidence sources and pipeline version used when it was generated.
Direction
The engine should grow without becoming a black box
New inputs and analyses should expand the project without removing the evidence trail that makes the results inspectable.
Near term
Additional DNA-provider formats, improved public beta tooling, stronger explanations and broader report content.
Longer term
Whole-genome sequencing, richer population and origins analysis, stronger pharmacogenomics and more interactive report experiences.
Throughout
Transparent evidence, visible limitations, cautious interpretation and user control over personal DNA.
Why it exists
Making raw DNA more useful without reducing it to a score
Raw consumer DNA files contain large numbers of genotype calls but are difficult to interpret directly.
Raw data is difficult to interpretConsumer files are not designed to be readable reports.
A typical consumer raw-data file contains hundreds of thousands of genotype calls. The file itself does not explain which observations are meaningful or how strong the evidence is.
Simple SNP lookups can remove contextA matching marker is not always enough for a useful conclusion.
DNA_CHECKS is intended to sit between raw genotype data and oversimplified one-line SNP interpretations: accessible enough to explore, but transparent enough to inspect.
Analysis model
The report is the end of a pipeline
DNA_CHECKS does not treat every scientific question as the same type of lookup.
01
Personal DNA
A supported raw genotype file supplied by the user.
02
Normalisation & QC
Variant identity and source-specific differences are handled before interpretation.
03
Evidence systems
Clinical, PGx, traits and origins are evaluated through specialised branches.
04
Report
Evidence is prioritised, contextualised and presented with its limitations.
Why keep evidence systems separate?Different scientific questions require different models.
Population frequency is not ancestry, modern population affinity is not ancient ancestry, pharmacogenomics is not disease diagnosis, and a trait association is not a clinical finding.
Why can evidence be withheld?The report does not need to be full of claims to be useful.
Evidence that fails coverage, quality or interpretation thresholds can be demoted, moved into supporting context or omitted.
What role can AI have?The underlying scientific result remains upstream and deterministic.
AI is not required to detect genotypes, classify findings or calculate ancestry comparisons. Where AI-assisted wording is used, its role is explanatory rather than scientific decision-making.
Explore
See the report itself
The public example shows the actual report structure without using personal DNA.