This project provides a Java implementation to convert your 23andMe genotype data into compressed VCFs (vcf.gz). VCF files can be useful for imputation with the Michigan Imputation Server or for mitochondrial haplogroup classification with HaploGrep2.
Your personal genome can be downloaded from here. After entering your secure answer, the complete dataset can be downloaded at once.
The VCF files are generated by combining information from your 23andMe data (.zip or .txt) with the human reference file. This project uses HTSJDK for many tasks.
git clone https://github.com/genepi/23andme-tools.git
cd 23andme-tools
mvn install
java -jar vcf-tools-0.1.jar vcf-generator --in </path/to/genome> --ref </path/to/human_g1k_v37.fasta>
--out <vcf-destination-folder> [--exclude <chromosomes to exclude>] [--split false]
The current version of the 23andMe genotyping chip v4 (> Nov 2013) is based on the GRCh37 reference (aka g1k, v37) human_g1k_v37.fasta. By setting --ref v37 it will be downloaded automatically.
This command downloads the fasta reference and writes each chromosome (chr1-22,X,Y,MT) to a seperate vcf.gz file:
java -jar vcf-tools-0.1.jar vcf-generator --in /path/to/genome
--ref v37 --out /path/to/vcfDir The chromosome MT file can be then used with e.g. HaploGrep2. Add --split false to write one VCF file.
The Imputation Server accepts vcf.gz files split by chromosome, which can be created with the following command:
java -jar vcf-tools-0.1.jar vcf-generator --in /path/to/23andMe-genome.txt
--ref /path/to/human_g1k_v37.fasta --out /path/to/vcfDir --exclude Y,MT