plink

Question

[ooo]

Hi all,

I use plink for gene association test.

command :

 plink --file myplink --noweb --allow-no-sex --geno 0.05 --mind 1 --hwe 0.001 --out cleaned

log :

 Options in effect:
        --file myplink
        --noweb
        --allow-no-sex
    --geno 0.05
    --mind 1
    --hwe 0.001
    --out cleaned

427635 (of 427635) markers to be included from [ myplink_ver3.map ]
Warning, found 621 individuals with ambiguous sex codes
Writing list of these individuals to [ cleaned_noAF_ver3.nosex ]
621 individuals read from [ myplink_ver3.ped ] 
0 individuals with nonmissing phenotypes
Assuming a disease phenotype (1=unaff, 2=aff, 0=miss)
Missing phenotype value is also -9
0 cases, 0 controls and 621 missing
0 males, 0 females, and 621 of unspecified sex
Before frequency and genotyping pruning, there are 427635 SNPs
621 founders and 0 non-founders found
95919 markers to be excluded based on HWE test ( p <= 0.001 )
    0 markers failed HWE test in cases
    0 markers failed HWE test in controls
Total genotyping rate in remaining individuals is nan
415605 SNPs failed missingness test ( GENO > 0.05 )
0 SNPs failed frequency test ( MAF < 0 )
After frequency and genotyping pruning, there are 7770 SNPs
After filtering, 0 cases, 0 controls and 621 missing
After filtering, 0 males, 0 females, and 621 of unspecified sex

I had this line in log file : Total genotyping rate in remaining individuals is nan.

Anyone has any idea on why this is?

Thank you

Answer 1

[LauferVA]

Without seeing your data, it is difficult to know if this is occurring due to problems with genotyping rate in the data or with problems dealing with the covariance matrix or some other calculation ... However, the number of SNPs failing HWE is way too high (95919), suggesting that your data themselves may be in need of further QC or processing ... I just don't know.

Alternatively, are you using plink 1.07? If so, please consider this thread.

In the thread, Chris Chang mentions that Plink 1.07 was handling singular covariance matrices incorrectly, while the (at that time Beta version, 1.9) was "less incorrect, but there's more work to be done. Will try to post a proper patch soon."

In any event what I would do first is break the command you have above into one step at a time, then closely look at the input and output files to figure out where the problems are ...