Problem on running HAWK 0.9.8 beta: on the EIG6.0.1-Hawk/bin/smartpca step

[Leandro Lima]

Hello,

I am trying to run HAWK 0.9.8 beta (downloaded from http://atifrahman.github.io/HAWK/) on bacterial strains where I do not have a reference genome. I managed to run HAWK 0.8.3 Beta on these strains, but I can't do it for the new HAWK version associated with the publication https://elifesciences.org/articles/32920. I am blocked when running EIG6.0.1-Hawk/bin/smartpca -V -p <hawk_dir>/parfile.txt. The main problem is that smartpca takes 3 input files:

genotypename: gwas_eigenstratX.geno
snpname: gwas_eigenstratX.snp
indivname: gwas_eigenstratX.ind

but the previous steps (k-mer counting with script supplements/countKmers and the lines before invoking smartpca in script supplements/runHawk) did not produce gwas_eigenstratX.geno and gwas_eigenstratX.snp, but only gwas_eigenstratX.ind. Since these files do not exist, smartpca fails, and since these 3 input files are required, I do not manage to run them. I also did not find any documentation on this regard.

Could you point me any direction on how to solve this?

Thank you in advance.

[atifrahman]

Did you update hawkDir correctly to run the latest version of hawk (0.9.8)? The new version should generate gwas_eigenstratX.geno and gwas_eigenstratX.snp. Lines 621 to 642 here
https://github.com/atifrahman/HAWK/blob/master/hawk.cpp

[Leandro Lima]

Hello,

Thanks for the fast and detailed answer. Hawk is running ok, the problem was on my end, I had a mistake in my runHawk shell script. Sorry about this, thank you again for the answer.