WFT?
Filed in: R
WFT? (For Wright-Fisher Tracker?) is a simple R function that does binomial sampling to show genetic drift.
Two functions: for folks not comfortable with programming, a simple for loop that lets you run a population of size N, with starting frequency p, for t generations. Prints out a table of each generation that can be pasted in excel to graph.
WF function: run as “WF(10,50,0.3)” to simulate a biallelic locus in a population of size N=50 (100 alleles) that starts at freq. p=0.3.
WF<-function(t,N,p){
print(c("gen","number of A's","freq. A (p)"))
for( g in 1:t ){
A=rbinom(1,N*2,p)
p=A/(2*N);
print(c(g, A, p),quote=F)
}
}
A slightly more sophisticated function that graphs X iterations.
WFgraf function: run as “WFgraf(10,50,0.3,10)” to simulate as above, but graphing 10 iterations.
WFgraf<-function(t,N,p,X){
freq=as.numeric();
gens=1:t
pnew=p
for( g in 1:t ){
A=rbinom(1,2*N,pnew)
pnew=A/(N*2);
freq[length(freq)+1]=pnew;
}
bob=data.frame(gens,freq)
plot(freq~gens,type="l",ylim=c(0,1),xlab="generations",ylab="frequency of A")
for( n in 2:X ){
freq=as.numeric();
gens=1:t
pnew=p
for( g in 1:t ){
A=rbinom(1,N*2,pnew)
pnew=A/(N*2);
freq[length(freq)+1]=pnew;
}
bob=data.frame(gens,freq)
lines(freq~gens,type="l",ylim=c(0,1),col=colors()[runif(1,1,657)])
}
}
An example graph using “WFgraf(100,100,.5,5)”
