Open Source

Your genome,
annotated.

Seven clinical databases. One pipeline. A self-contained HTML report you can open anywhere. No cloud. No account. No data leaves your machine.

View on GitHub See sample report →
$ pip install allelix Click to copy
Annotation Sources
Seven databases. One pass.
Each variant is cross-referenced against clinical, functional, and population databases. All annotations land in a single report.
C
ClinVar
NCBI's archive of variant-condition relationships with clinical review status.
Clinical significance
P
PharmGKB
Pharmacogenomic associations linking genotypes to drug response phenotypes.
Drug interactions
S
SNPedia
Community-curated wiki of SNP effects, magnitudes, and genotype interpretations.
Magnitude scoring
G
GWAS Catalog
EBI/NHGRI catalog of published genome-wide association studies.
Trait associations
A
AlphaMissense
DeepMind's pathogenicity predictions for all possible single amino acid substitutions.
ML pathogenicity
D
CADD
Combined Annotation Dependent Depletion scores for variant deleteriousness.
Phred-scaled scores
n
gnomAD
Population allele frequencies from 807,162 individuals across diverse ancestries.
Population frequency
Architecture
Designed for privacy.
Your genotyping data never leaves your machine. The entire pipeline runs locally.
🔒
Fully offline
No network requests at runtime. Reference databases are downloaded once. Reports open via file:// protocol.
📄
Single-file reports
Self-contained HTML with inline CSS and JS. Email it, archive it, open it in any browser. No server required.
🔬
Multi-format input
23andMe, AncestryDNA, MyHeritage, and other DTC genotyping formats. PLINK export for research workflows.
⚖️
Composite scoring
Cross-source magnitude system inspired by SNPedia. Highest severity across all databases surfaces first.
💻
Pure Python
No compiled extensions. pip install and run. Works on macOS, Linux, and Windows with Python 3.10+.
🛠️
Open source
AGPL-3.0 licensed. Every annotation decision is traceable to source. No black-box algorithms.
Coming soon VCF & whole-genome sequencing support in v2.0
Output
The report.
Interactive, sortable, searchable. Click any variant to see all source annotations in a detail sidebar.
file:///Users/you/allelix_report.html
23andMe
Format
GRCh37
Build
184
Variants
23
Bad
47
Good
Search variants...
All Bad Good Neutral ☾
Mag Gene Geno Repute Summary
8 BRCA2 AG Bad [ClinVar +1] Breast/ovarian cancer susceptibility
6 APOE CT Bad [SNPedia +2] Alzheimer's disease risk factor
4 CYP2D6 GG Good [PharmGKB] Normal metabolizer phenotype
3 MTHFR CC Good [ClinVar] Normal folate metabolism
2 COMT AG Neutral [GWAS +1] Common variant, no clinical significance

Annotate your genome in minutes.

Annotate your raw genotyping file. Get a fully annotated, self-contained report. Keep it forever.

Get started Documentation