covariates issue

[CuteGold0407]

Hi, sorry to disturb, recently I have been running the cis_nominal mode, with covariates files with different number of rows. These rows contains 1 row for gender, 5 for PCA of genotype, and 0/10/20/30 for peerfactors of phenotype. But using different peerfactors of phenotype output the same cis-nominal mapping result. so I wonder why and how to fix it? thanks! this is my shell command: python3 -m tensorqtl \ xxx.genotype \ xxx.pheno.bed.gz \ nominal_PEER${i}_000001 \ #${i} means covariates files with different number (0/10/20/30) of peerfactors --covariates cov/covariates_PEER${i}.txt.gz \ --mode cis_nominal \ --window 1000000 \ --maf_threshold 0.00001

[CuteGold0407]

actually, when i cancel the -- covariates argument, the command still runs with results just like the before, so i guess maybe i didn't add the --cov argument correctly?

[CuteGold0407]

I go back to check again and found the --covariates was added correctly and the format of the cov file also seems right, cov/geno/pheno all have the paired unique ID, so why this problem happens(;_;)

[francois-a]

Can you provide the output from the log? You should see the line * X covariates, and that number should change depending on your input. Your MAF threshold seems too low.

[CuteGold0407]

Thank you! it's the maf problem!