Post date: Jan 26, 2015 6:3:1 PM
I am calculating the genomic autocorrelation (Moran's I) in Fst for each pair of population's in Moe's data set. I couldn't do this for all loci (too much memory), so for now I am just using the first 100,000 SNPs. I can always repeat this with another set of SNPs. The infiles and results for this are in /labs/evolution/projects/timema_radiation/popgen/fsts/autoc/. Also, I modified the genomac.C file slightly to automatically center the data. First, I split the Fst values by population pair (from a file with all Fsts, of which I took the first 100,000):
tail -n +2 sub100k_fst.txt | cut -f 1 -d " " > sub100k_BCBOG.txt
tail -n +2 sub100k_fst.txt | cut -f 2 -d " " > sub100k_BCTUR.txt
tail -n +2 sub100k_fst.txt | cut -f 3 -d " " > sub100k_BMCG3.txt
tail -n +2 sub100k_fst.txt | cut -f 4 -d " " > sub100k_BMT.txt
tail -n +2 sub100k_fst.txt | cut -f 5 -d " " > sub100k_BS.txt
tail -n +2 sub100k_fst.txt | cut -f 6 -d " " > sub100k_CR.txt
tail -n +2 sub100k_fst.txt | cut -f 7 -d " " > sub100k_HF.txt
tail -n +2 sub100k_fst.txt | cut -f 8 -d " " > sub100k_LP.txt
tail -n +2 sub100k_fst.txt | cut -f 9 -d " " > sub100k_SM.txt
tail -n +2 sub100k_fst.txt | cut -f 10 -d " " > sub100k_VP.txt
Then I ran the genomic autocorrelation analysis (via a perl submission wrapper; here are two examples).
cd /labs/evolution/projects/timema_radiation/popgen/fsts/autoc/
~/bin/genomac -n 100000 -c subCoords.txt -p sub100k_SM.txt -l bins.txt -o fstAc_sub100k_SM.txt
cd /labs/evolution/projects/timema_radiation/popgen/fsts/autoc/
~/bin/genomac -n 100000 -c subCoords.txt -p sub100k_VP.txt -l bins.txt -o fstAc_sub100k_VP.txt