Generate vcf.gz file and its index file vcf.gz.tbi

Question

[lyz10302012]

Can anyone tell me how to generate vcf.gz file and its index file vcf.gz.tbi in 1000 Genomes Project?

ftp://ftp-trace.ncbi.nih.gov/1000genomes/ftp/release/20110521/

Answer 1

[Matt Shirley]

bgzip -c file.vcf > file.vcf.gz
tabix -p vcf file.vcf.gz

tabix documentation

[Dan]

Does the VCF have to be sorted like SAM/BAM does?

[Matt Shirley]

Yes, tabix requires sorted input files. I don't think the sorting order matters, but records must be grouped together by rows.

[Dan]

You mean numeric or alphanumeric by chromosome in ascending or descending order plus numeric by position ascending or descending order? I can't think how else sorting order could not matter.

[Matt Shirley]

Yes. To clarify, I think you just need your chromosomes grouped together, and then records for each chromosome need to be sorted in ascending coordinate order.

Answer 2

[Erik Garrison]

I have a script which does this using a VCF stream on stdin:

#!/bin/bash

file=$1

bgzip >$file
tabix -f -p vcf $file

I found I was always writing the same lines over and over when indexing VCF files. You can use it like this:

cat uncompressed.vcf | bgziptabix compressed.vcf.gz

Answer 3

[miaowzai]

bgzip file.vcf       # or:   bcftools view file.vcf -Oz -o file.vcf.gz
tabix file.vcf.gz    # or:   bcftools index file.vcf.gz

this is convenient where tabix and bgzip are not installed. saw this from: https://github.com/samtools/bcftools/issues/668

Answer 4

[ATCG]

bgzip genotypes.vcf && tabix -p vcf genotypes.vcf.gz

see details if you need to:

https://qtltools.github.io/qtltools/pages/input_files.html