Plotting Fst

Question

[thjnant]

Hello,

I have a huge SNP dataset of 25 millions SNP genome wide. I have divided it into different chromosomes and this FST plot is for chromosome 1. To generate this plot and for it to be visually tractable, I have chosen SNPs every 50 bp. But the plot is still not very good and there are many many SNPs with the same overall FST value. I was wondering is there any way to reduce the noise in this plot, kind of taking down most points, something like logarithm. Any help is very much appreciated.

Thanks a lot in advance.

[Pavel Senin]

Seems like you are using ggplot to plot these, did you try to use jitter and add alpha value to your dots, so they will be semi-transparent?

Answer 1

[Pavel Senin]

Here, I think that jitter helps with that line of constant values

require(ggplot2)
require(Cairo)
require(grid)
data = data.frame(x=c(runif(15000)*1e+08,runif(450)*1e+08),
                  y=c(runif(15000),rep(0.3,450)))
p <- ggplot(data, aes(x, y)) 
p1 <- p + geom_point(cex=0.9)
p2 <- p + geom_jitter(position = position_jitter(w = 0.01, h = 0.02), 
                      colour="darkorchid3", alpha=0.3,cex=0.9)
Cairo(width = 800, height = 400, file="test.png")
print(grid.arrange(p1, p2, ncol=2))
dev.off()