Fastest Way To Rename Sample Name In 37 Gb Gzipped Vcf File Or Binary Ped File

Question

[Kevin]

Hi I have a project file for 2000 individuals, with SNPs from WGS. We have decided to change the sample names to a more uniform way of naming individuals from different sources.

I wish to do a quick check if there's inbuilt functions in plink or plink seq or vcftools to do this i.e. pass it a sample name remapping file to output a new set of fam ids and idv ids.

Otherwise, i might consider doing this using sqlite since there's a possibility I might have to slice the data in another way again

[Davy]

Plink can update individual names in FAM and ped files and the like.

http://pngu.mgh.harvard.edu/~purcell/plink/dataman.shtml#updatefam

however I know from PLINKSEQ just changing the files, might not actually work.

[stevenlang123]

Hi, running into an identical problem here. Do you recall how you ended up doing this?

Thanks in advance,
Steve

Answer 1

[tiagoantao]

As long as you do NOT (i) change the order or (ii) remove/add individuals you can edit the FAM file for ids. My recommendation would be a script to do the correction on the FAM file.

Remember: the order and the number of individuals stay the same. Other than that, you can change the FAM ids

Answer 2

[Nick Crawford]

You could create a new vcf header with the final #CHROM line updated to reflect the new sample names. Then use tabix to replace the older header with the new one.

# Replace VCF header. The file must be compressed by bgzip. 
tabix -r header.txt big.vcf.bgz > big.vcf.new_header.bgz

More here: http://vcftools.sourceforge.net/docs.html#one-liners

Edit: To get your VCF in the correct format and to create the tbi index do something like...

gunzip -c big.vcf.gz | bgzip > big.vcf.bgz
tabix -p vcf big.vcf.bgz