Post date: Oct 16, 2013 3:11:8 PM
We decided to include a brief analysis of the exceptional change loci on adenostoma from the transplant (within-generation) experiment in response to reviewer comments. I am a bit unsure whether we have 171 or 173 of these, so I am using 171. Basically, I just summarized selection (differential) coefficients for these loci, and made a plot. Here is what I found,
(i) Summary of S for the 171 loci with P(exceptional) >= 0.95:
mean = 0.248
sd = 0.0438
min = 0.0785
max = 0.344
2.5th q = 0.135
97.5th q = 0.312
(ii) S is significantly different from 0 for 94 of these loci (94/171 = 55%)
These results fit nicely with the rest of the story. Even though we had more exceptional change loci on Adenostoma, fewer of these have significant values for S, and the average S is lower (in fact the maximum S on Adenostoma is similar to the mean on Ceanothus).
These results and relevant files are in projects/timema_mortality_experiment/. The R code I used is here:
nloci<-186576
psel<-scan("pselTrtA.txt",n=nloci,sep=" ")
S<-read.table("outsTrtAwigs.txt",header=TRUE)
X<-which(psel >= 0.975)[1:171]
mean(abs(S$median[X]))
## 0.2407747
sd(abs(S$median[X]))
## 0.04384199
summary(abs(S$median[X]))
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.07852 0.22250 0.24760 0.24080 0.26820 0.34380
pdf("sTrtA.pdf",width=8,height=10)
par(mar=c(4.5,4.5,0.5,0.5))
plot(S$median[X],1:171,xlab="selection coefficient",ylab="locus",pch=20,type='n',cex.lab=1.5,las=2,cex.axis=1.3,xlim=c(-0.55,0.55))
segments(S$ci0.025[X],1:171,S$ci0.975[X],1:171,lwd=0.9,col="darkgray")
points(S$median[X],1:171,pch=20)
abline(v=0,col="red",lwd=1.2,lty=2)
dev.off()
sum(S$ci0.025[X] > 0 | S$ci0.975[X] < 0)
## 94
## this is number significant S of 171