how to adjust a Manhattan plot for bayes factor

Hi all, I am trying to draw a Manhattan plot for the output of my genome-scan analysis. this is the head of my data-set (I have bayes factor for each enviornmental variable; BF_env4_p instead of p-value, ranging from -30 to +50).

head(chrom1)
        snp_pos chrom BF_env4_p   BF_env5_p   BF_env6_p  snp
    1   52967     1  -5.892840  -5.445123  -6.327080 52967_0
    2   53241     1 -11.409509 -11.438922 -11.109112 53241_0
    3   68062     1  -6.409155  -5.544443  -6.774634 68062_0
    4   68080     1 -11.501740 -12.907424  -8.474243 68080_0
    5   68285     1 -12.400456 -12.633168 -12.630723 68285_0
    6   68296     1 -11.319704 -10.951429 -12.062912 68296_0

This the code that I am using

manhattan(chrom1,bp= "snp_pos", chr="chrom",p="BF_env4_p",snp = "snp", ylim=c(-30,50))

but the plot is really strange, and range of dots is only around "0". Is there any way that I could fix it for BF instead of p-value? in addition I get an error message that says

In manhattan(chrom1, bp = "snp_pos", chr = "chrom", p = "env4_p",  :
  NaNs produced

any suggestion is appreciated.

manhattan plot bayes factor

• 2,862 views • link

updated 8.7 years ago by Biostar

• written 8.8 years ago by Ana

OK, I got the solution after reading "Manhattan" , I changed my code to this:

> manhattan(chrom1,bp= "snp_pos", chr="chrom",p="env4_p",snp = "snp",
 logp = FALSE, ylim=c(-20,50))

The problem was logp, If TRUE, the -log10 of the p-value is plotted. It isn't very useful to plot raw p-values, but plotting the raw value could be useful for other genome-wide plots, for example, peak heights, bayes factors, test statistics, other "scores," etc.

• 0 views • link

written 8.8 years ago by Ana

0 answers

No answers yet.