pQTL Mediation analysis

Description Usage Arguments Examples

Triad plot. Currently relies on sdp to provide lines, but want to use coefficients from model fit with mediation_test to get lines for each column of driver. Note that the plot uses column info to provide additional information, which here is the chr of mediator. The plot uses the mediator position on its home chromosome, which is not really what is wanted. See package qtl2shiny for a more elegant use.

mediation_triad(target, mediator, driver, covar_tar = NULL,
  covar_med = NULL, kinship = NULL, fitFunction = fitQtl2, ...)

ggplot_mediation_triad(x, tname = "target", mname = "mediator",
  dname = "driver", centerline = NULL, fitline = FALSE,
  main = paste(tname, "by", mname, "and", dname),
  colors = qtl2::CCcolors, ...)

## S3 method for class 'mediation_triad'
autoplot(object, ...)

## S3 method for class 'mediation_triad'
plot(x, ...)

`target`	vector or 1-column matrix with target values
`mediator`	vector or 1-column matrix with mediator values
`driver`	vector or matrix with driver values
`covar_tar`	optional covariates for target
`covar_med`	optional covariates for mediator
`kinship`	optional kinship matrix among individuals
`fitFunction`	function to fit models with driver, target and mediator
`...`	additional arguments to internal routine
`x`	object of class `mediation_triad`
`tname`	target name (default `"target"`)
`mname`	mediator name (default `"mediator"`)
`dname`	driver name (default `"driver"`)
`centerline`	horizontal line at value (default = `NULL`); set to `NA` for no line or `NULL` for mean
`fitline`	include fit line from coefficients in `x` if `TRUE`
`main`	main title (defautl `tname`)
`colors`	named colors to use if `fitline` is `TRUE`
`sdp`	SNP distribution pattern for plot colors
`allele`	Driver has alleles if `TRUE`, otherwise allele pairs.
`...`	additional parameters for plotting

data(Tmem68)
target <- cbind(Tmem68 = Tmem68$target)
# Pick strongest mediator that is not target.
m <- match("Nnt", Tmem68$annotation$symbol)
mediator <- Tmem68$mediator[, m, drop = FALSE]
colnames(mediator) <- "Nnt"
# Reconstruct 8-allele genotype probabilities.
driver <- cbind(A = 1 - apply(Tmem68$qtl.geno, 1, sum), Tmem68$qtl.geno)
rownames(driver) <- rownames(Tmem68$qtl.geno)

med_triad <- mediation_triad(target = target,
                      mediator = mediator,
                      driver = driver,
                      covar_tar = Tmem68$covar,
                      sdp = 2)
ggplot2::autoplot(med_triad, tname = "Tmem68", mname = "Nnt")