meanScan: Make a scanone plot of genotype means.

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

Description

meanScan Like qtl:effectscan, but plots the genotypic means across the marker / pseudomarker grid.

Usage

meanScan(cross, pheno.col = 1, chr = NULL, covar = NULL,
  mean.FUN = function(x) mean(x, na.rm = T), se.FUN = function(x) sd(x,
  na.rm = T)/sqrt(sum(!is.na(x))), sw.width = 1, plot.se = T,
  se.alpha = 0.2, col = NULL, leg.pos = "topright", ...)

Arguments

cross	The qtl cross. Must contain genotype probabilities or imputed genotypes
pheno.col	Character or numeric vector indicating the phenotype to be tested. Only 1 phenotype can be tested at a time. If numeric, converted to character.
chr	The chromosomes to analyze. If not supplied, all are tested.
covar	dataframe of covariates like addcovar in qtl::scanone. Can only be a single column. If the data is numeric, or character, converted to factor first.
mean.FUN	The function applied to generate the line positions. Defaults to mean(x, na.rm = T)
se.FUN	The function applied to generate the confidence interval around the mean. Defaults to sd(x, na.rm = T)/sqrt(sum(!is.na(x)))
sw.width	The window size to use if smoothing is desired. If 1, no sliding window smoothing is conducted. Must be smaller than the number of markers on the smallest chromosome.
plot.se	Logical, should standard errors around the menas be plotted?
se.alpha	Numeric coding of the transparency for the standard error regions.
col	Character vector of colors to plot.
leg.pos	The character of numeric (length 2) position of the legend. Default is "topright".
...	additional arguments passed on to plot.scanone
leg.pos	The position of the legend.

Details

Calculates genotypic means at each marker/psuedomarker, plots them.

Value

The plot and an object of class scanone containing the means for each combination of genotype and covariate.

Examples

## Not run: 
library(qtlTools)
library(qtl)
data(fake.bc)
cross<-fake.bc
cross<-calc.genoprob(cross)
ms<-meanScan(cross, covar = NULL)

# Some simple customization
ms<-meanScan(cross, covar = NULL, col = c("black","orange"),
   leg.pos = "top")

sex<-data.frame(sex = ifelse(pull.pheno(cross, "sex") == 0,"F","M"))
ms<-meanScan(cross, covar = sex, leg.pos = "right")
ms<-meanScan(cross, covar = sex, chr = c(2,7), leg.pos = "right",
   col = c("black","grey","darkblue","lightblue"))

cross.fem<-subset(cross, ind = sex[,1] == "F")
cross.male<-subset(cross, ind = sex[,1] == "M")
par(mfrow = c(2,1))
ms <-meanScan(cross.fem, leg.pos = "topright", leg.inset = .05, main = "females only")
ms <-meanScan(cross.male, leg.pos = "topright", leg.inset = .05, main = "males only")

data(fake.f2)
cross<-fake.f2
cross<-calc.genoprob(cross)
ms <- meanScan(cross, covar = NULL, leg.pos = "bottomright", leg.inset = .1)
sex<-data.frame(sex = ifelse(pull.pheno(cross, "sex") == 0,"F","M"))
ms<-meanScan(cross, covar = sex)

data(fake.4way)
cross<-fake.4way
cross<-calc.genoprob(cross)
ms<-meanScan(cross, covar = NULL)
sex<-data.frame(sex = ifelse(pull.pheno(cross, "sex") == 0,"F","M"))
ms<-meanScan(cross, covar = sex, chr = c(2,7))

data("multitrait")
cross =multitrait
cross<-calc.genoprob(cross)
ms<-meanScan(cross, covar = NULL)
ms<-meanScan(cross, covar = NULL, sw.width = 5)

## End(Not run)

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