This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Plink Alternative Phenotype File Columns not being Read

Hi, I have a plink alternative phenotype file with the following format:

FID IID Phenotype
1    2        1
1    3        0
..

As outlined in the plink documentation https://zzz.bwh.harvard.edu/plink/data.shtml#pheno

However, when I run the following command :

plink --bfile ../Plink_Files/plink  --logistic --adjust  --pheno ../Pheno_Files/pheno.txt --all-pheno --covar ../Cov_Files/pheno.txt   --covar-number 2-4  --allow-no-sex --out Results/pheno

I get the following error message:

4768 people (0 males, 0 females, 4768 ambiguous) loaded from .fam.
Ambiguous sex IDs written to Results/Fibroids.nosex .
397 phenotype values present after --pheno.
Using 1 thread (no multithreaded calculations invoked).
--covar: 3 out of 11 covariates loaded.
1271 people were not seen in the covariate file.
Before main variant filters, 4768 founders and 0 nonfounders present.
Calculating allele frequencies... done.
Total genotyping rate is 0.943943.
94036 variants and 4768 people pass filters and QC.
Among remaining phenotypes, 0 are cases and 397 are controls.  (4371 phenotypes
are missing.)
397 phenotype values present after --pheno.
Pheno has 0 cases, 397 controls, and 4371 missing phenotypes.

Which implies that the phenotype file is being read where the rows are phenotypes not the columns because it says the phenotypes are missing. I am not sure why that would be since the phenotype file matches the plink format that I found online. Any help in fixing this error would be greatly appreciated!

plink1.9
• 1,743 views • link
updated by Ram Ram • written by dtlloyd • dtlloyd

1 answer

There could have two reason. For your case，high possibility is second reason.

In the phenotype file some values could be missing e.g. -9 or NA values

      FID  IID  Phenotype
       1     1    1
       2     2    1
       3     3    2
       4     4    -9

As you included the covariate file and used covariate 2-4 (means 3 covariates) in the model, there may also have missing in the covariate corresponding to the individuals.

    FID   C1   C2   C3 
    1    12    31   NA 
    2    17    NA   25 
    3    NA    20   15
• 0 views • link
written by asif_stat_raj • asif_stat_raj

Log in to answer this question.

More posts like this