VCF To 23andMe Converter

Converts VCF files into 23andMe raw data format.

CLI

genomics convert vcf-to-23andme --config configs/converters/vcf_to_23andme/default.yaml

Code And Configs

Kind	Path
Code	src/genomics/converters/vcf_to_23andme/
Configs	configs/converters/vcf_to_23andme/

Features

• automatic rsID annotation with dbSNP through BCFtools;
• GRCh37/GRCh38 auto-detection from VCF metadata;
• V3 and V5 output formats;
• optional chip-panel filtering;
• multi-sample VCF support;
• read-only handling of the input VCF.

Implementation Structure

The converter is implemented primarily in converter.py and exposed through __main__.py for CLI delegation.

Key responsibilities in converter.py:

Area	Implementation detail
Build detection	Compares contig metadata and chromosome lengths against GRCh37/GRCh38 constants
dbSNP support	Downloads/caches dbSNP VCFs and indexes for GRCh37 or GRCh38
Annotation	Uses BCFtools to annotate missing rsIDs when configured
Chromosome normalization	Handles chr1, 1, RefSeq accessions, X/Y/MT variants
Chip panels	Downloads/caches Illumina panel manifests and extracts rsIDs
Output formatting	Emits 23andMe-like V3/V5 headers and tab-separated SNP records

Conversion Flow

input VCF
  -> detect sample and genome build
  -> optionally annotate rsIDs with dbSNP
  -> optionally filter to chip panel or custom SNP panel
  -> normalize chromosome names and genotypes
  -> write 23andMe raw text file

Config Sections

Section	Purpose
input	Input VCF path and optional sample name
output	Output path, format version, rsID/chip-panel behavior
annotation	dbSNP annotation and reference cache directory
reference_build	Fallback build when auto-detection is inconclusive

External Requirements

dbSNP annotation requires bcftools on PATH. Use source scripts/env/start_genomics_universal.sh before running conversion in the project environment.