Post date: Jun 27, 2016 2:1:31 AM
Maximum likelihood inference of allele frequencies for the pilot fluctuating selection-diversity-time series data set. Now working in /uufs/chpc.utah.edu/common/home/u6000989/projects/lycaeides_diversity/.
1. Create a genotype likelihood file from the morefilter*vcf file. Here I am retaining SNVs with MAF > 1% (i.e., dropping rare variants). This might or might not be a good idea. I will revisit it later. This leaves me with 20,775 variants. The results are in lyc_timeseries.gl.
perl splitPops.pl lyc_timeseries.gl
2. Now I am splitting on population and year.
perl splitPops.pl lyc_timeseries.gl
3. Filter out sex linked SNVs and those without known LGs (only work with autosomal SNVs).
perl retainAutos.pl lyctimeseries_*gl
4. ML inference of allele freqs. with estpEM. Results are in /uufs/chpc.utah.edu/common/home/u6000989/projects/lycaeides_diversity/AlleleFreqs/.
perl ../wrap_qsub_slurm_estpem.pl auto_lyctimeseries_*
cd /uufs/chpc.utah.edu/common/home/u6000989/projects/lycaeides_diversity/AlleleFreqs
~/bin/estpEM -i auto_lyctimeseries_USL-15.gl -o outp_USL-15.txt -m 50 -h 2