Phenotype from plink fam file in gcta

Question

[bha]

My question is about specifying phenotype file in GCTA analysis. My phenotype data is in my .fam file(sixth column). But software notes says i should make another test.phen. is it possible that GCTA read phenotype from .fam file?

# MLMA analysis - If you have already computed the GRM
gcta64 --mlma --bfile test --grm test --pheno test.phen --out test --thread-num 10

I tried to look into gcta forum, but could get answer, does anyone came across this? any suggestion please?

Answer 1

[Kevin Blighe]

It looks like it should be possible to create the phen file manually. Here is what is required:

[source: http://cnsgenomics.com/software/gcta/GCTA_UserManual_v1.24.pdf]

So, you just need the FID, IID, and Phenotype. If you have your original FAM file, then you just need to do:

cut -f1,2,6 test.fam > test.phen

Also, if your starting point is a PLINK dataset, you can export it to GCTA GRM format with --make-grm-gz / --make-grm-bin

Kevin

[bha]

Many thanks Kevin, but basic point was that if phenotype is already in test.fam, then what is point to have separate --pheno test.phen. Anyways thanks for responding

[Kevin Blighe]

Probably a question more for the developers of GCTA, in that case. It is a program design issue. If there is no indication in the GCTA manual that a FAM file is used, then, of course, one should not assume that the FAM is used.

[seta]

Hi Kevin,

In the case of a longitudinal cohort study, the SNP association analysis can be done by the linear mixed model as implemented in MLMA, yes? Could you please tell me how to put all measurements of a given variable (trait) at each time point in the "test.phen" file?

Many thanks