I am trying to run an association analysis with plink version 1.07 using dosage files. I am working with impute2 output files but I managed to replicate the problem in a small artificial dataset.
I created a 10 individual 5 snp dataset and ran the commands
plink --fam test.fam --dosage test.dat format=3 dose1 noheader --linear --out No_covar
and
plink --fam test.fam --dosage test.dat format=3 dose1 noheader --linear --covar covar --out With_covar
I expected the second command to give me information regarding the covariates, as it does for association analysis on ped/map files. Instead, for both cases I got a file on the same form with only one line per SNP and no TEST column. Is that a plink bug? Or is it just the way covariate analysis works for dosage files?
The results I obtained were, without covariates:
SNP A1 A2 FRQ INFO OR SE P snp1 A T 0.6500 1.3407 0.0000 18234.1705 0.9984 snp2 G C 0.7500 0.6667 5.9930 1.8006 0.32 snp3 T A 1.0000 0.0000 NA NA NA snp4 T G 1.0000 0.0000 NA NA NA snp5 C A 0.5000 2.0000 0.0000 9039.7038 0.9981
and with covariates.
SNP A1 A2 FRQ INFO OR SE P snp1 A T 0.6500 1.3407 0.0000 10993.8403 0.9971 snp2 G C 0.7500 0.6667 2.2500 1.2910 0.5299 snp3 T A 1.0000 0.0000 NA NA NA snp4 T G 1.0000 0.0000 NA NA NA snp5 C A 0.5000 2.0000 0.0000 7709.0853 0.9978
The test.fam file:
1 1 0 0 1 1 2 2 0 0 2 1 3 3 0 0 1 1 4 4 0 0 2 1 5 5 0 0 1 1 6 6 0 0 2 2 7 7 0 0 1 2 8 8 0 0 2 2 9 9 0 0 1 2 0 0 0 0 2 2
the test.dat file
snp1 A T 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0 1 0 0 0 1 0 0 1 snp2 G C 0 1 0 1 0 0 0 1 0 1 0 0 0 1 0 1 0 0 0 1 0 1 0 0 0 1 0 1 0 0 snp3 T A 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 snp4 T G 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 snp5 C A 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1
and the covar file
1 1 1 2 2 2 3 3 3 4 4 4 5 5 5 6 6 6 7 7 7 8 8 8 9 9 9 0 0 0