manhatton.plot: Manhattan plot of p-values

In plantbreeding: Analysis and visualization of data from plant breeding and genetics experiments

Description

The function develops Mahnattan plot of p-values scaled to -log10(p). If polar type of Manhattan plot is desired use the function manhatton.circos. Manhattan plot (Gibson 2010) are popular in plotting association mapping results, however can be used to plot other results genome-wide.

Usage

manhatton.plot(dataframe, SNPname, chromosome, position, pvcol, ylabel = "pvalue", pconv= "-log10", ymax = "maximum",
 ymin = "minimum", gapbp = 500, pch = c(18, 19, 20), color = c("midnightblue", 
 "lightpink4", "blue"), line1, line2)

Arguments

dataframe	dataframe with SNP name (SNPname), chromosome, physical position (position), p-value columns.
SNPname	Name of variable consisting of SNP name - (eg. "SNPN")
chromosome	Name of variable column consisting of chromosome - ( eg. "chr")
position	Name of variable column consisting of physical position of SNPs - (eg. "physicaldis")
pvcol	Name of p-value column to be used for plotting, dataframe can consists of multiple p-value column, can be plotted one by one. Note that p-value should not contain zero or Inf or NaNs
ylabel	Y axis label, only if pconv is not "-log10", defualt is "pvalue"
ymax	Maximum value to be plotted in Y axis, if ymax is less than 8, the plot will set the maximum to 8 otherwise user defined maximum.
ymin	Minimum value to be plotted in X axis.
gapbp	Gap between two adjecent chromsomome for plotting. Should be specified to scale of distances provided for X axis (ie. base pair). The default value is 500.
pch	The list of symbol type used to plot in the plot, maximum allowed is equal to number of chrosomomes plotted. If the number is less than total number of chromosomes, the pch is recycled till end.
color	The list of color type used to plot in the plot, maximum allowed is equal to number of chrosomomes plotted. If the number is less than total number of chromosomes, the color is recycled till end. The number of color should be equal to number of pch.
line1	Value at the point where you need to Horizental threshold line 1. NULL for no line
line2	Value at the point where you need to Horizental threshold line 2. NULL for no line

Details

Most of plot prameters (not conflicting with specified here) can be applied to plot.

Value

Produce Manhattan plot

Author(s)

Umesh Rosyara

Examples

# Example 1
data12 <- data.frame (snp = 1: 2000*20 , chr = c(rep(1:20, each = 2000)), 
pos= rep(1:2000, 20), pval= rnorm(2000*20, 0.001, 0.005))

manhatton.plot(dataframe = data12, SNPname = "snp", chromosome = "chr", 
position = "pos", pvcol = "pval",ymax = "maximum", ymin = 0,  gapbp = 500, 
color=c("hotpink3","dodgerblue4"), line1 = 3, line2 = 5, pch = c(1,20) )

manhatton.plot(dataframe = data12, SNPname = "snp", chromosome = "chr", position = "pos",
 pvcol = "pval",ymax = 10, ymin = 2,  gapbp = 500, color=c("dodgerblue4"), 
 line1 = 3, line2 = 5, pch = 20 )
   
manhatton.plot(dataframe = data12, SNPname = "snp", chromosome = "chr", position = "pos", 
pvcol = "pval",ymax = "maximum", ymin = 0,  gapbp = 500, 
color=c("midnightblue", "lightpink4", "blue"),
 line1 = 3, line2 = 5, pch = c("*", "+", "a") )
 
manhatton.plot(dataframe = data12, SNPname = "snp", chromosome = "chr", position = "pos", 
pvcol = "pval",ymax = "maximum", ymin = 0,  gapbp = 500, color= "cadetblue", 
line1 = 3, line2 = 5, pch = 19)

# all different color and pch example 
cbPalette <- c("#999999", "#E69F00", "#56B4E9", "#009E73", 
"#F0E442", "#0072B2", "#D55E00", "#CC79A7","#CD661D", "#FF00FF","#8B6508",
 "#D2691E","#008B00", "#8B1A1A",  "#8B3A62", "#8B864E", "#3CB371", "#8B5742",
  "#8B5A00", "#36648B")
                                                                                                                                                                       
manhatton.plot(dataframe = data12, SNPname = "snp", chromosome = "chr", position = "pos",
 pvcol = "pval",ymax = "maximum", ymin = 0,  gapbp = 500, color= cbPalette, 
 line1 = 3, line2 = 5, pch = 1:20)

# Example 2
set.seed(123)
data22 <- data.frame (snp = 1: 20000*5 , chr = c(rep(1:5, each = 20000)), 
pos= rep(1:20000, 5), pval1= rnorm(20000*5, 0.2, 0.3), 
pval2 = rnorm(20000*5, 0.2, 0.3) )
# the above simulation produce negative values so the following will replace 
# negative values with NA
data22$pval1[data22$pval1 < 0] <- NA
# removal of negative values 
dat2 <- data22[!is.na(data22$pval1),]
 
op <- par(mfrow=c(2,1), cex.axis = 0.75, font = 1, family = "serif")
par(op)
manhatton.plot(dataframe = dat2, SNPname = "snp", chromosome = "chr", position = "pos",
 pvcol = "pval1", line1 = 4, line2 = 8, ymax = "maximum", ymin = 0, gapbp = 2000)
title(main = "Mahattan plot of results for trait1", sub = "Method: Linear mixed model")

manhatton.plot(dataframe = dat2, SNPname = "snp", chromosome = "chr", position = "pos",
 pvcol = "pval2", line1 = 4, line2 = 8, ymax = "maximum", ymin = 0)
title(main = "Mahattan plot of results for trait2", sub = "Method: Linear mixed model")

# Example 3, plotting simple p-value or R-square 
 data13 <- data.frame (snp = 1: 2000*20 , chr = c(rep(1:20, each = 2000)), 
pos= rep(1:2000, 20), pval= rnorm(2000*20, 0.15, 0.05))
manhatton.plot(dataframe = data13, pconv= "NULL", ylabel = "R-square", SNPname = "snp", chromosome = "chr", 
position = "pos", pvcol = "pval",ymax = "maximum", ymin = 0,  gapbp = 500, 
color=c("hotpink3","dodgerblue4"), line1 = 3, line2 = 5, pch = c(1,20) )

plantbreeding documentation built on May 2, 2019, 4:54 p.m.