Post date: Jun 28, 2016 5:34:56 PM
Here is a comparison using the Bayesian estimates of allele frequencies from popmod. I decided to use these as they give more similar allele frequencies (which suggests they are more accurate), though the correlation between the two methods is very very high (> 0.99).
perl ../wrap_qsub_slurm_popmod.pl auto_lyctimeseries_*
cd /uufs/chpc.utah.edu/common/home/u6000989/projects/lycaeides_diversity/AlleleFreqs
~/bin/popmod -i auto_lyctimeseries_USL-15.gl -o outp_bayes_USL-15.hdf5 -b 1000 -n 10000 -t 5 -a 0.5
cut -f 2 -d "," ../AlleleFreqs/bp_BCR-13.txt | perl -p -i -e 's/^(\S+)/\1 98/' > in_pb_BCR-13.txt
cut -f 2 -d "," ../AlleleFreqs/bp_BCR-15.txt | perl -p -i -e 's/^(\S+)/\1 82/' > in_pb_BCR-15.txt
varne -a in_pb_BCR-13.txt -b in_pb_BCR-15.txt -l 10096 -t 2 -n 3000 -x 1000
hat{Ne} = 153.861
median of posterior = 154.639
50% credible intervals = 144.747, 164.556
90% credible intervals = 133.79, 180.169
95% credible intervals = 129.789, 187.169
cut -f 2 -d "," ../AlleleFreqs/bp_BNP-13.txt | perl -p -i -e 's/^(\S+)/\1 98/' > in_pb_BNP-13.txt
cut -f 2 -d "," ../AlleleFreqs/bp_BNP-15.txt | perl -p -i -e 's/^(\S+)/\1 98/' > in_pb_BNP-15.txt
varne -a in_pb_BNP-13.txt -b in_pb_BNP-15.txt -l 10096 -t 2 -n 3000 -x 1000
hat{Ne} = 107.639
median of posterior = 108.107
50% credible intervals = 102.699, 113.729
90% credible intervals = 95.7675, 121.721
95% credible intervals = 94.0908, 125.47
cut -f 2 -d "," ../AlleleFreqs/bp_BTB-13.txt | perl -p -i -e 's/^(\S+)/\1 98/' > in_pb_BTB-13.txt
cut -f 2 -d "," ../AlleleFreqs/bp_BTB-15.txt | perl -p -i -e 's/^(\S+)/\1 102/' > in_pb_BTB-15.txt
varne -a in_pb_BTB-13.txt -b in_pb_BTB-15.txt -l 10096 -t 2 -n 3000 -x 1000
hat{Ne} = 160.754
median of posterior = 160.814
50% credible intervals = 152.185, 172.469
90% credible intervals = 139.199, 189.378
95% credible intervals = 135.572, 194.928
cut -f 2 -d "," ../AlleleFreqs/bp_GNP-13.txt | perl -p -i -e 's/^(\S+)/\1 98/' > in_pb_GNP-13.txt
cut -f 2 -d "," ../AlleleFreqs/bp_GNP-15.txt | perl -p -i -e 's/^(\S+)/\1 102/' > in_pb_GNP-15.txt
varne -a in_pb_GNP-13.txt -b in_pb_GNP-15.txt -l 10096 -t 2 -n 3000 -x 1000
hat{Ne} = 146.321
median of posterior = 146.989
50% credible intervals = 137.874, 156.024
90% credible intervals = 125.693, 172.627
95% credible intervals = 122.973, 179.702
cut -f 2 -d "," ../AlleleFreqs/bp_MRF-13.txt | perl -p -i -e 's/^(\S+)/\1 96/' > in_pb_MRF-13.txt
cut -f 2 -d "," ../AlleleFreqs/bp_MRF-15.txt | perl -p -i -e 's/^(\S+)/\1 100/' > in_pb_MRF-15.txt
varne -a in_pb_MRF-13.txt -b in_pb_MRF-15.txt -l 10096 -t 2 -n 3000 -x 1000
hat{Ne} = 124.103
median of posterior = 124.207
50% credible intervals = 116.951, 130.672
90% credible intervals = 108.25, 142.788
95% credible intervals = 105.568, 145.697
cut -f 2 -d "," ../AlleleFreqs/bp_PSP-13.txt | perl -p -i -e 's/^(\S+)/\1 98/' > in_pb_PSP-13.txt
cut -f 2 -d "," ../AlleleFreqs/bp_PSP-15.txt | perl -p -i -e 's/^(\S+)/\1 76/' > in_pb_PSP-15.txt
varne -a in_pb_PSP-13.txt -b in_pb_PSP-15.txt -l 10096 -t 2 -n 3000 -x 1000
hat{Ne} = 125.299
median of posterior = 125.663
50% credible intervals = 117.954, 133.537
90% credible intervals = 108.455, 146.752
95% credible intervals = 105.254, 151.064
cut -f 2 -d "," ../AlleleFreqs/bp_RNV-13.txt | perl -p -i -e 's/^(\S+)/\1 98/' > in_pb_RNV-13.txt
cut -f 2 -d "," ../AlleleFreqs/bp_RNV-14.txt | perl -p -i -e 's/^(\S+)/\1 94/' > in_pb_RNV-14.txt
varne -a in_pb_RNV-13.txt -b in_pb_RNV-14.txt -l 10096 -t 1 -n 3000 -x 1000
hat{Ne} = 84.5333
median of posterior = 85.5423
50% credible intervals = 79.7633, 91.1554
90% credible intervals = 72.6239, 101.491
95% credible intervals = 69.7343, 103.917
cut -f 2 -d "," ../AlleleFreqs/bp_USL-13.txt | perl -p -i -e 's/^(\S+)/\1 98/' > in_pb_USL-13.txt
cut -f 2 -d "," ../AlleleFreqs/bp_USL-15.txt | perl -p -i -e 's/^(\S+)/\1 100/' > in_pb_USL-15.txt
varne -a in_pb_USL-13.txt -b in_pb_USL-15.txt -l 10096 -t 2 -n 3000 -x 1000
hat{Ne} = 143.711
median of posterior = 144.846
50% credible intervals = 136.624, 153.616
90% credible intervals = 126.16, 166.041
95% credible intervals = 123.514, 171.607
A few summaries of the results:
Plot of variance effective population size (point estimate and 90% CIs):
Variance Ne:
mean = 131 (2Ne - 263)
range = 85-161 (2Ne = 170-322)
cv = 0.196, range/mean = 57.8\%
Comparison with expected expected het. (which exhibits substantially less variation)
het. range = 0.105-0.109
cv = 0.014, range/mean = 0.0396
No correlation was detected between heterozygosity and Ne:
Pearson's product-moment correlation
data: het and ne$bayes
t = -0.1997, df = 6, p-value = 0.8483
alternative hypothesis: true correlation is not equal to 0
95 percent confidence interval:
-0.7433732 0.6612705
sample estimates:
cor
-0.08127487