Post date: Feb 08, 2017 3:51:9 AM
The ABC model from the 2016 Evolution paper was updated for the L14 line. The code is in cvs and the new version is compiled on kingspeak as abc-dyn. I obtained the set of 224 SNPs that deviated from the null model from Alex. The file pLentilSnps.txt is in /uufs/chpc.utah.edu/common/home/u6000989/scratch/wfabc/. I ran the ABC model with perfect additivity or incomplete dominance for each SNP. The outfiles are in /uufs/chpc.utah.edu/common/home/u6000989/scratch/wfabc/ and have the SNP number as well as 'additive' or 'incomp' for the run type. The job was executed as sbatch abc.sh, which runs forkRunAbc.pl. The lattern contains the following (note that this runs 5 million ABC iterations per SNP and model):
#!/usr/bin/perl
#
# creates a child process per 25 SNPs
#
#$dir = "/uufs/chpc.utah.edu/common/home/u6000989/scratch/wfabc/";
foreach $proc (0..9){
$pid = fork;
if ($pid) { ## fork successful
$forks++;
$lb = 1 + $proc * 25;
if($proc < 9){
$ub = $lb + 25;
}
else{
$ub = 224;
}
for($i=$lb; $i<$ub; $i++){
system "wfabc-dyn -i pLentilSnps.txt -l $i -n 5000000 -v 0.1 -h 0 > out_additive_snp$i.txt\n";
system "wfabc-dyn -i pLentilSnps.txt -l $i -n 5000000 -v 0.1 -h 1 > out_incomp_snp$i.txt\n";
}
exit;
}
}
for (1 .. $forks){
$pid = wait();
print "Parent saw $pid exit\n";
}