Error with log_reg_case

[Asutu]

Hi,

I'm running into an error when calling the log_reg_case.out program.

$ $hawkDir/log_reg_case.out -t $noThread $sex_confounder_arg -p $noPC > pvals_case_top.txt
terminate called after throwing an instance of 'alglib::ap_error'
Aborted

the values of the variables set above are: noThread=1, noPC=2, and sex_confounder_arg is not set.

The number of lines in the (truncated) output is much less than those in the original sorted files:

$ wc -l pvals_case_top.txt
790000 pvals_case_top.txt
$ wc -l case_out_w_bonf_top.kmerDiff
107759269 case_out_w_bonf_top.kmerDiff

Any ideas on how to fix this?
Many thanks,
Pedro

[atifrahman]

Hi Pedro,

Sorry about this!

This sort of error may happen if some input file is not provided or provided empty or incomplete (e.g. pcs.evec, incomplete gwas_info.txt). Maybe checking the input files will reveal something,

Would it be possible to share the following files? We can take a look
pcs.evec
gwas_info.txt
gwas_eigenstratX.ind
case_total_kmers.txt
control_total_kmers.txt
case_out_w_bonf_top.kmerDiff
control_out_w_bonf_top.kmerDiff

[Asutu]

Of course, many thanks for looking into this.

See the first 5 files attached. I had to gzip gwas_eigenstratX.ind and pcs.evec to be able to upload them here.
However, both case_out_w_bonf_top.kmerDiff and control_out_w_bonf_top.kmerDiff are way too big to share this way (9.6G and 3.7G, respectively).

case_total_kmers.txt
control_total_kmers.txt
gwas_eigenstratX.ind.gz
gwas_info.txt
pcs.evec.gz

[atifrahman]

One of the developers had a look and didn't find any problem in the files you provided. The problem may be with the kmerDiff files. If you can share those via Google Drive or something like that, we can try to find a solution. My email is [email protected]

Sorry about the problems