Post date: Jan 05, 2016 10:6:47 PM
The binary runs with the probit model gave many PVEs near 0 or 1. I suspect this might reflect poor mixing. Thus, I am re-running these with the linear model to see what effect this has.
Here is an example command (data set 1):
perl ../wrap_qsub_rc_gemma.pl /labs/evolution/projects/selection_sims/geno_sims.txt /labs/evolution/projects/selection_sims/phenobin_sims1.txt
And here is an example execution command:
cd /pscratch/A01963476/sims/gemmasel
sleep 5
gemma -g /labs/evolution/projects/selection_sims/geno_sims.txt -p /labs/evolution/projects/selection_sims/phenobin_sims1.txt -bslmm 1 -n 50 -o out_sim1binarylin50rep1 -rpace 100 -maf 0.0 -w 1000000 -s 2000000