v-pagano/ancestry-snpweights

Ancestry estimation by SNPweights

There are many software tools available to estimate ancestry using genetic data. The method "SNPweights" Chen et al. 2013 Bioinformatics uses SNP weights precomputed from large external reference panels. It presented a unique approach to leverage the rich ancestry information that is available in WGS or WES data.

The software package "Ancestry-SNPweights" is a wrapper that integrated varies software tools to ease the usage of the software 'SNPweights', and enables users to take genetic data in popular formats such as vcf, bam/cram, plink and Illumina studio report format of genotyping arrays.

Software dependencies

python 3.6
pyvcf (depends on python 3.6)
bcftools
vcfutils.pl from bcftools package
samtools
Picard tools
plink

Notes: if you don't take array data, pyvcf and python3.6 are not required.

Software installation

git clone https://github.com/tgen/Ancestry-SNPweights

cd Ancestry-SNPweights/data

unzip snpwt.NA.zip

cat snpwt.AS.gz.0* > snpwt.AS.gz 

gunzip snpwt.AS.gz

Usage examples

1. VCF input in Hg38/GRCh38 for a single sample

python ~/compute/github/Ancestry-SNPweights/infer_ancestry_vcf.py \
    -v /path/to/vcf_file \
    -o /output/dirctory

2. Bam or cram input in Hg38/GRCh38 for a single sample

python ~/compute/github/Ancestry-SNPweights/infer_ancestry_vcf.py \
    -b /path/to/bam_file \
    -o /output/dirctory

3. Array genotype report input in GRch37 for a single sample

python ~/compute/github/Ancestry-SNPweights/infer_ancestry_vcf.py \
    -v /path/to/Illumina_genome_studio_report_file \
    -o /output/dirctory

4. Plink input data for a cohort of samples

To be implemented.

Output examples

input file name:  <file_name_prefix>.vcf.gz
output file name: <file_name_prefix>.predpc.csv

Outputs:

sample_1:
AFR,EUR,SAS,EAS,NAT,Ancestry
0.145,0.816,0.0,0.039,.,EUR

sample_2:
AFR,EUR,SAS,EAS,NAT,Ancestry
0.853,0.147,0,0,.,AFR

sample_3:
AFR,EUR,SAS,EAS,NAT,Ancestry
0.0,0.014,.,0.06,0.925,NAT

Sample_4
AFR,EUR,SAS,EAS,NAT,Ancestry
0.074,0.502,.,0,0.424,Mix(EUR+NAT)

Sample_5
AFR,EUR,SAS,EAS,NAT,Ancestry
0.35,0.479,.,0.171,0,Mix(EUR+AFR)

Column keys:

AFR: Affican
EUR: Europian
SAS: South Asian
EAS: East Asian
NAT: Native American

Ancestry: Assigned population (>=0.8)