storeylab / gcat

This package has been renamed gcatest: https://github.com/StoreyLab/gcatest

This package has been renamed gcatest. The github is here: https://github.com/StoreyLab/gcatest

gcat's Issues

Handling NA data

Sorry this is not the right way to raise issues - I haven't used github before.

1. I am trying to use GCAT and when I applies lfa to my genotype data it reports

Error in if ((S < SVTol) && (j == 1)) stop("error: starting vector near the null space") : missing value where TRUE/FALSE needed

Which I presume to be due to the missingness in the data (genotype data is cleaned in PLINK to geno and mind 0.03). However in your paper you report cleaning the North Finland Cohort data at 1% missingness. Can the gcat method handle missingness in PLINK genotype data, and if so how should the missingness be coded?

2. The GCAT method appears to be very memory intensive for even small sets (1000 people, 700K SNPs) - do you know roughly how it scales in terms of RAM needed for N indivduals and M SNPs?

Thanks

Matthew