Extract all genotype counts from phased data in vcf files

Question

[spiral01]

I have vcf files (1000 genomes) with phased data and would like to extract counts of the genotypes. Several methods were suggested here: Genotype count in a VCF file

However, I need to extract not just the homozygous for ref, heterozygous, and homozygous for alt counts, but the counts for every combination. I.e. I need counts of:

0|0 0|1 1|0 1|1

for each site. Is this possible?

Answer 1

[Pierre Lindenbaum]

using bioalcidaejdk: http://lindenb.github.io/jvarkit/BioAlcidaeJdk.html

$ wget -q -O - "http://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/ALL.chr22.phase3_shapeit2_mvncall_integrated_v5a.20130502.genotypes.vcf.gz" |\
gunzip -c |\
java -jar dist/bioalcidaejdk.jar -F VCF -e 'stream().forEach(V->println(V.getContig()+" "+V.getStart()+" " +V.getReference().getDisplayString()+" "+V.getGenotypes().stream().map(G->G.getAlleles().stream().map(A->String.valueOf(V.getAlleleIndex(A))).collect(Collectors.joining(G.isPhased()?"|":"/"))).collect(Collectors.groupingBy(Function.identity(), Collectors.counting()))));'

22 16050075 A {0|0=2503, 0|1=1}
22 16050115 G {0|0=2472, 1|0=16, 0|1=16}
22 16050213 C {0|0=2467, 1|0=18, 0|1=18, 1|1=1}
22 16050319 C {0|0=2503, 1|0=1}
22 16050527 C {0|0=2503, 0|1=1}
22 16050568 C {0|0=2502, 0|1=2}
22 16050607 G {0|0=2499, 1|0=3, 0|1=2}
22 16050627 G {0|0=2502, 1|0=2}
22 16050646 G {0|0=2503, 1|0=1}
22 16050654 A {0|0=1834, 1|0=3, 0|1=6, 0|2=48, 2|0=37, 0|3=255, 3|0=258, 4|0=6, 0|4=12, 2|3=1, 3|2=1, 3|3=41, 4|3=2}
22 16050655 G {0|0=2503, 1|0=1}
22 16050678 C {0|0=2502, 0|1=2}
22 16050679 G {0|0=2503, 1|0=1}
22 16050688 C {0|0=2503, 1|0=1}
22 16050732 C {0|0=2503, 0|1=1}
22 16050739 TA {0|0=2467, 1|0=18, 0|1=18, 1|1=1}
22 16050758 T {0|0=2503, 0|1=1}
22 16050783 A {0|0=2466, 0|1=13, 1|0=24, 1|1=1}
22 16050840 C {0|0=2478, 1|0=11, 0|1=15}
22 16050847 T {0|0=2498, 1|0=2, 0|1=4}