segmentsOnMap: Visualize the positions of QTL on a genetic map.

In jtlovell/qtlTools: Data Processing and Plotting in Association with R/qtl

Description

segmentsOnMap A basic function to plot QTL confidence intervals. Useful for <20 traits.

Usage

segmentsOnMap(cross, phe, chr, l, h, peaklod = NA, peakcM = NA,
  calcCisResults = NULL, legendPosition = "bottom", legendCex = 0.8,
  col = NULL, palette = highContrastColors, lwd = "byLod",
  leg.lwd = 2, max.lwd = 5, min.lwd = 1, tick.width = NULL,
  leg.inset = 0.01, chrBuffer = c(0.05, 0.15), orderBy = "lod",
  showPeaks = FALSE, ...)

Arguments

cross	The qtl cross object with marker names that need to be changed.
phe	Character or numeric vector indicating the phenotype to be tested
chr	Vector of chromosome ids - will be coerced to numeric
l	The lower confidence interval bound for each qtl
h	The upper confidence interval bound for each qtl
peaklod	Optional - if provided, permits segments width to be weighted by significance.
peakcM	Optional - the position of the peak
calcCisResults	A shortcut that allows results from calcCis to be piped directly into the plotting function. If provided, l, h, chr, phe and peaklod are ignored
legendPosition	Where to place the legend. This is passed to the first argument of the legend function. See legend documentation
legendCex	The character expansion for the legend. This is passed to the cex argument of legend.
col	Optional. A vector of colors for segments
palette	Optional. A color palette with which to draw segment colors
lwd	Either the specification "byLod", where segment weights are calculated from LOD score data, or a numeric vector of length 1 that is passed to segments lwd
leg.lwd	The lineweights of the legend.
max.lwd	If scaling by LOD, what should the maximum lwd be?
min.lwd	If scaling by LOD, what should the minimum lwd be?
tick.width	How wide should the ticks be? Defaults to 1/n chromosomes.
leg.inset	How far should the legend be away from the plot border?
chrBuffer	Numeric vector of length 2, specifying the buffer space between confidence interval segments on the left and right adjacent chromosomes.
orderBy	What data should the segments be stacked by?
showPeaks	Logical, should the QTL peaks be plotted as points?
...	Other arguments passed to segments

Details

Pass output from bayesint, lodint, or another confidence interval estimation program to visualize this.

Value

The plot

Examples

library(qtlTools)
data(multitrait)
segmentsOnMap(cross = multitrait, phe = paste0("phe",1:15),
 chr = rep(1,15), l =seq(from = 1, to = 100, length.out = 15),
 h =seq(from = 20, to = 120, length.out = 15))
segmentsOnMap(cross = multitrait, phe = paste0("phe",1:25),
 chr = rep(1,25), l =seq(from = 1, to = 100, length.out = 25),
 h =seq(from = 20, to = 120, length.out = 25))
## Not run: 
# use multitrait data
cross <- multitrait
cross <- calc.genoprob(cross)
phes <- phenames(cross)[1:20]
# Conduct scanone and permutations to get confidence intervals
s1 <- scanone(cross, pheno.col = phes, method = "hk")
perms <- scanone(cross, pheno.col = phes, method = "hk",
   n.perm = 100, verbose=F)
cis<-calcCis(cross, s1.output = s1, perm.output = perms)

# manual construction of the confidence intervals
with(cis, segmentsOnMap(cross, phe = pheno, chr = chr,
   l = lowposition, h = highposition, legendCex = .5,
   peakcM = pos,
   tick.width = .1,  chrBuffer = c(.15,.2)))
with(cis, segmentsOnMap(cross, phe = pheno, chr = chr,
   l = lowposition, h = highposition, legendCex = .5,
   tick.width = .1,  chrBuffer = c(.15,.2)))

# feed calcCis directly into the plot
segmentsOnMap(cross, calcCisResults = cis, legendCex = .5)
segmentsOnMap(cross, calcCisResults = cis, legendCex = .5,
   col = terrain.colors(length(unique(cis$phe))))
segmentsOnMap(cross, calcCisResults = cis, legendCex = .5,
   max.lwd=6, min.lwd=.5)
segmentsOnMap(cross, calcCisResults = cis, legendCex = .5,
   max.lwd=4, min.lwd=2)
segmentsOnMap(cross, calcCisResults = cis, legendCex = .5,
   lwd = 2)
segmentsOnMap(cross, calcCisResults = cis, legendCex = .5,
   palette = rainbow)

## End(Not run)

jtlovell/qtlTools documentation built on May 20, 2019, 3:14 a.m.