Best pipeline to merge multiple consumer raw DNA files into PLINK + EIGENSTRAT (1240K) with minimal SNP loss?
Hi all,
I have consumer “raw DNA” genotype text files for the same individual from multiple services:
AncestryDNA, 23andMe, FTDNA, MyHeritage, LivingDNA, MapmyGenome, tellmeGen and GenoConnect.
My goal is to produce one merged dataset in both PLINK (bed/bim/fam) and EIGENSTRAT (geno/ind/snp), restricted to the 1240K SNP set, while keeping SNP loss minimal.
Is there a recommended methodology or pipeline available to complete this?
Any pointers to an existing pipeline, scripts, or a methodology would be appreciated.
Thank you.
0 answers
No answers yet.
Log in to answer this question.
More posts like this
I don't have any expertise here, but have you looked around on github ? Most bioinfo software is hosted there these days and I'm sure I have seen related packages in the past.
Hi colindaven,
I have had a look around Github. However I was just wondering if there's a standardised accepted pipeline/methodology.
I'm not aware of a single tool/pipeline that converts all the mentioned formats into plink. As the formats and reference genome versions across providers vary, you may need to write multiple scripts to make them compatible with plink.
Some pointers:
https://github.com/castedo/genomology
https://www.jade-cheng.com/au/23andme-to-plink/
https://www.rpubs.com/tiffamee/phd251d-project2
https://github.com/teepean/PLINK-to-23andMe
Hi Arup,
Thank you for the links.
Yeah I agree there are many differing formats, I'll have a look into these ones linked.
Thank you
What format are the files in? Proprietary for each service or something standard like VCF?
Hi GenoMax,
Vast majority are from Ancestry and 23andme. A sizeable amount are from FTDNA too.
They’re raw genotype text exports (tab-delimited rsID/chr/pos plus genotype/alleles), not VCF. Usually downloaded as .txt/.zip but not a standard VCF.