This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Conversion tools for GWAS

Good morning, actually I'm working for a GWAS for the metabolic syndrome, I've .bed .bim and .fam files also .ped and .map, I'm looking for a tool that allow me to obtain .geno file , because I need to do some research with EIGENSOFT and it needs some file that I can't have right now. There is some tool that is capable to convert some of my file into .geno file.

I need also .ind and .snp and .pca files how can i obtain these files also?

(I apologize for my English I'm not British or American)

genome-wide gwas

But i can't understand how it works, can you give me an example of command-line rule?

convertf -p params.txt where params.txt looks something like:

genotypename: mydata.bed
snpname: mydata.map
indivname: mydata.fam
outputformat: PACKEDANCESTRYMAP
genotypeoutname: converted.geno
snpoutname: converted.snp
indivoutname: converted.ind
familynames: NO

Thank you for your reply. May I ask how can I convert .bed, .bim and .fam files to packedancestrymap format?

Can I also use converf? If so, how should my parameter file look like?

Many thanks!

@joey I will try when possible and i will tell you , thank you so much

@joey It gives me this error:

sh: 1: smartpca: not found
OOPS couldn't open file MDC_OMXEX_zcall_QC_pass.pca.evec for reading at ./smartpca.perl line 65.

Have you checked the PATH?

Nope I didn't put that now I will try it,

smartpca: error while loading shared libraries: liblapack.so.3: cannot open shared object file: No such file or directory
OOPS couldn't open file MDC_OMXEX_zcall_QC_pass.pca.evec for reading at ./smartpca.perl line 65.

Those are the errors, I can't figure it out

You need to update your system or download some libraries.

For example in Ubuntu you can install libatlas package using

sudo apt-get install libatlas3-base

which should have the so.3 library

I'm so sorry and I need a little more help smartpca:

smartpca: /usr/lib/x86_64-linux-gnu/libgfortran.so.1: version `GFORTRAN_1.4' not found (required by /usr/lib/liblapack.so.3)
smartpca: /usr/lib/x86_64-linux-gnu/libgfortran.so.1: version `GFORTRAN_1.4' not found (required by /usr/lib/libblas.so.3)

(I'm on Ubuntu 14.04)

So..I fixed the problem at last...

I looked at the Makefile and saw that it was calling gfortran

FC = gfortran

which wasn't there on my file system...

I change it to FC = gfortran-4.4 and recompiled the porgram using

make clobber
make install

and it's working fine.

Thank you all, thank you Joey

hi,danielestrife.

I have suffered the same problems with you. I'm glad that you have conquered them, but i don't understand how you solve it at the final process as you descripted in this post. Could you give me some help? Thank you!

2 answers

You can still work with the PLINK binary files:

cp study.bim study.pedsnp ## Renaming so that Eigensoft can recognize file type
cp study.fam study.pedind ## Renaming so that Eigensoft can recognize file type

smartpca.perl -i study.bed -a study.pedsnp -b study.pedind -o study.pca -p study.plot -e study.eval -m 0 -l study.log

rm study.pedsnp ## removing temp files
rm study.pedsnp ## removing temp files

The CONVERTF program in EIGENSOFT handles this.

Unfortunately, this function only supports chromosome numbers 1-23. I work in non-model systems, and have a draft genome with several thousand scaffolds. This is not a problem in plink with the --allow-extra-chr option, but EIGENSOFT doesn't have a way around this to convert those .PED and .MAP files to the .GENO file that I need.

Log in to answer this question.