confects_plot_me: Mean-expression vs effect size plot (deprecated)

In pfh/topconfects: Top Confident Effect Sizes

Mean-expression vs effect size plot (deprecated)

Description

Note: I now recommend using plot_confects_me2 instead of this plot. Like plotMD in limma, plots effect size against mean expression level. However shows "confect" on the y axis rather than "effect" ("effect" is shown underneath in grey). This may be useful for assessing whether effects are only being detected in highly expressed genes.

Usage

confects_plot_me(confects)

Arguments

confects

A "Topconfects" class object, as returned from limma_confects, edger_confects, or deseq2_confects.

Value

The two types of points in this plot make it quite confusing to explain. plot_confects_me2 is recommended instead.

A ggplot2 object. Working non-interactively, you must print() this for it to be displayed.

Examples

library(NBPSeq)
library(edgeR)
library(limma)

data(arab)

# Extract experimental design from sample names
treat <- factor(substring(colnames(arab),1,4), levels=c("mock","hrcc"))
time <- factor(substring(colnames(arab),5,5))

# Keep genes with at least 3 samples having an RPM of more than 2
y <- DGEList(arab)
keep <- rowSums(cpm(y)>2) >= 3
y <- y[keep,,keep.lib.sizes=FALSE]
y <- calcNormFactors(y)

# Find top confident fold changes by topconfects-limma-voom method
design <- model.matrix(~time+treat)
voomed <- voom(y, design)
fit <- lmFit(voomed, design)
confects <- limma_confects(fit, "treathrcc")

# Plot confident effect size against mean expression
# (estimated effect size also shown as grey dots)
confects_plot_me(confects)

pfh/topconfects documentation built on Dec. 12, 2024, 10:19 p.m.