How Mutant Works

Mutant turns supported DNA data into structured, educational health context. It maps selected genetic variants into biological modules, organizes those modules into systems and focused hubs, identifies converging patterns, and surfaces ranked health hypotheses with the supporting evidence and limitations clearly shown.

The result is not a diagnosis or a claim that Mutant has found the cause of a health problem. It is a more organized way to explore what your genetics may suggest—and how those findings may fit alongside the rest of your health information.

From Your DNA File to an Educational Analysis

Step 1: Add a Supported DNA File

Create a free Mutant account and select the raw DNA .txt file from 23andMe or AncestryDNA, or a supported whole-genome .vcf or .vcf.gz file. If your provider supplies SNPs and insertions/deletions separately, you can select both files.

Mutant normally processes the raw file locally in your browser and uploads the selected genetic variants needed to create your analysis. Those variants and your analysis are saved with your account.

The current upload flow can upload your full DNA file if your browser cannot process it locally, or if a whole-genome file yields no matching variants. Local processing therefore does not apply to every upload.

Keep the DNA page open while processing completes. Whole-genome files can take a few minutes.

Step 2: Relevant Variants and Coverage Are Identified

Mutant checks your file for the variants included in its analysis and records which are available or missing.

Coverage depends on the source and format of your DNA data. Microarray files such as 23andMe and AncestryDNA provide useful coverage of many common variants, while supported whole-genome files generally provide broader coverage.

Your analysis displays its coverage, and missing data is not treated as a normal result.

Step 3: Variants Are Organized Into Biological Modules

Rather than presenting a long list of isolated SNPs, Mutant groups related variants into biological modules.

A module represents a focused biological process, such as signaling, transport, enzyme activity, immune regulation, nutrient handling, or cellular response. Each module is evaluated using the relevant genetic signals, their direction, evidence, and relative importance.

Modules can contain several small-effect signals or, in some cases, a more influential variant. Mutant preserves that distinction so users can see what is supporting each result.

How Evo 2 Adjusts Variant Weights

Evo 2 is a DNA sequence model developed by Arc Institute. It learns patterns in DNA sequences rather than in variant-by-variant association data.

Mutant uses Evo 2 as a weighting step inside its existing scoring, not as a separate source of evidence:

Catalog variant → precomputed Evo 2 score → adjusted SNP weight → module or matched storm score

Mutant scores the variants in its catalog ahead of report generation, using the DNA sequence surrounding each catalog variant. That precomputed result adjusts the weight the variant already had when Mutant evaluates a module or matches a multi-variant pattern. A variant without a usable Evo 2 score keeps its original weight.

A difference in model likelihood is not a measure of clinical harm, and it is not a personal disease-risk percentage. Evo 2 changes how strongly an existing variant contributes; it does not add a separate evidence score, diagnose a condition, or score every variant.

See the Evo 2 FAQ answers for how this appears in a report.

Step 4: Modules Are Mapped Into Systems and Focused Hubs

Modules are organized into broad biological systems and their related hubs:

This structure makes it easier to explore related findings together and see when the same genetic tendency may be relevant across more than one system.

Step 5: Weighted Modules Form Converging Patterns

Mutant looks for themes supported by multiple biological modules.

Each module contributes according to the strength and relevance of its underlying signals. When several weighted modules point toward an overlapping biological theme, Mutant identifies a converging pattern.

Converging patterns are not diagnoses and do not prove that a biological process is impaired. They show where multiple pieces of genetic evidence support a related question more strongly than an isolated variant would on its own.

Step 6: Mutant Ranks Health Hypotheses

Mutant uses the module and pattern evidence to surface health hypotheses that may be worth exploring further.

Each hypothesis includes:

• A plain-language explanation of what the genetic pattern may suggest
• The genetic evidence and any contributing modules or converging patterns behind the ranking
• Supporting variants and evidence quality
• Genetic coverage and important missing data
• Relevant limitations and uncertainty
• Potential clinical correlations that may help guide further investigation

A higher ranking indicates greater priority within your analysis based on the available genetic evidence. It is not a probability that a condition is present, and it does not confirm a hypothesis.

Selected Clinical & Inherited Health findings require variant confirmation and interpretation alongside relevant laboratory results, medical history, and family history.

Connect Genetic Findings With Your Health History

DNA is only one part of a health picture. Symptoms, laboratory results, medications, medical history, environment, and lifestyle may support, challenge, or change the relevance of a genetic hypothesis.

Mutant provides the genetic analysis. You choose which additional health information to bring into an AI conversation or a discussion with a qualified healthcare professional.

Review Your Genetic Findings in Mutant

Explore your systems, hubs, ranked health hypotheses, and supporting genetic evidence directly in Mutant. Coverage and limitations help you understand what your DNA data supports and which findings need further investigation.

When Using Mutant Through ChatGPT

If you use Mutant in ChatGPT, connect your Mutant account and review the genetic-data processing and result-sharing permissions during that connection's setup. These permissions apply to the ChatGPT connection.

Mutant can then return relevant findings when you ask about them, within your account's Free or Full access. No export is required for this workflow.

You remain in control of the information you choose to provide or connect in ChatGPT. Mutant supplies the requested genetic findings; it does not automatically receive your full conversation, medical records, or connected health data.

You can disconnect Mutant from the ChatGPT connection permissions screen. This stops future access through that connection; it does not remove findings already shared in a conversation.

When Using Another AI Tool

Select Export for AI to download a structured, portable Health Context. You can then combine it with the records, laboratory results, medications, symptoms, and other information you choose to provide to a compatible AI tool.

Evidence and Limitations Remain Visible

Genetic findings differ in evidence quality, effect size, and relevance. Mutant is designed to show those differences rather than present every signal with the same certainty.

Keep these principles in mind when reviewing an analysis:

• Genetic susceptibility is not certainty.
• Association does not establish causation.
• Missing data is not a normal result.
• A hypothesis is not a diagnosis.
• DNA cannot be interpreted responsibly without considering other health information.
• AI-generated explanations can be useful, but they can also be incomplete or incorrect.

Mutant gives you an educational framework for further exploration. Medical concerns and decisions should be discussed with a qualified healthcare professional.

What You Can Do With Your Results

Inside Mutant, you can:

• Move between biological systems and focused hubs
• Review your highest-ranked health hypotheses
• Examine the modules and patterns supporting each hypothesis
• Inspect relevant variants, evidence quality, coverage, and limitations
• Retrieve your findings through the Mutant ChatGPT plugin
• Export structured context for use with another compatible AI tool

For the clearest sense of how this works, explore the sample just as you would your own analysis.

Start With the DNA Data You Already Have

Mutant Free includes your top three ranked health hypotheses in full, their evidence breakdowns, and the related findings available within your Free access.

Mutant Full unlocks all ranked health hypotheses in your analysis, their full evidence breakdowns, contributing patterns, and supporting variants for $49/year, billed annually. See Free and Full access details.

You can begin with supported microarray data and add supported whole-genome data later if you want broader genetic coverage.

After signing in, account controls let you clear a report's DNA data and derived analysis or request account closure. Deletion removes Mutant-stored data; it does not remove copies you already exported or shared with another service. Read the Privacy Policy and Data Policy for details.

No credit card required to create a free account.

Mutant Genomics is for educational and informational purposes only. It does not provide medical advice, diagnosis, or treatment. Always consult a qualified healthcare professional about medical concerns and decisions.