metatab.show: Display meta-analysis results.

In metamicrobiomeR: Microbiome Data Analysis & Meta-Analysis with GAMLSS-BEZI & Random Effects

Description

This function displays meta-analysis results of relative abundance as heatmap, forest plot, table or data.

Usage

metatab.show(
  metatab,
  com.pooled.tab,
  sumvar = "taxa",
  highest.lev = "g",
  tax.lev = "l2",
  showvar,
  estimate.pattern = "Estimate.",
  se.pattern = "Std. Error.",
  p.pattern = "Pr(>|t|)",
  readjust.p = FALSE,
  p.cutoff.type = "p",
  p.cutoff = 0.05,
  display = "plot",
  plot = "heatmap",
  fill.value = "log(OR)",
  grid = FALSE,
  digit = 2,
  p.digit = 4
)

Arguments

metatab	matrice of taxa/pathway abundance comparison meta-analysis results generated from meta.taxa.
com.pooled.tab	matrice of taxa/pathway abundance comparison generated from taxa.compare or pathway.compare combined from all included studies.
sumvar	Either "taxa" for taxa and "path" for pathway.
highest.lev	Highest level of bacterial taxonomies available for analysis. Options are "g" for genus (usually for 16S data) and "s" for species (usually for shortgun data).
tax.lev	taxa level to be displayed. Options are from "l2" (phylum) to "l7" (species). Default is "l2".
showvar	variable (string pattern) in the model to be displayed.
estimate.pattern	string pattern for estimates. Default is "Estimate.".
se.pattern	string pattern for standard error variable. Default is "Std. Error.".
p.pattern	string pattern for p-value variable. Default is "Pr(>|t|)".
readjust.p	multiple testing re-adjustment for only the level to be displayed (TRUE) or keep original multiple testing adjustment for all taxa of all levels (FALSE).Default is FALSE.
p.cutoff.type	type of p-value for cutoff. Options are "p" for p-value or "p.adjust" for multiple testing adjusted p-value. Default is "p".
p.cutoff	cutoff p-value to be displayed. Default is 0.05.
display	type of display. Options are display=c("plot","table","data")
plot	type of plot. Options are plot=c("heatmap","forest").
fill.value	name of legend.
grid	whether multiple plots will be displayed alongside. Default is FALSE.
digit	digit for estimates and 95 CI. Default is 2.
p.digit	digit for p-values. Default is 4.

Value

plot table or data.

Examples

# Load saved GAMLSS-BEZI results of four studies for the comparison of
# bacterial taxa relative abundance between genders adjusted for
# breastfeeding and infant age at sample collection
data(tabsex4)
#select only taxonomies of a small phylum for meta-analysis example
# (to save running time)
tlm<-tabsex4$id[grep("k__bacteria.p__fusobacteria",tabsex4$id)]
# meta-analysis
metab.sex<-meta.taxa(taxcomdat=tabsex4[tabsex4$id %in% tlm,],
summary.measure="RR", pool.var="id", studylab="study",
backtransform=FALSE, percent.meta=0.5, p.adjust.method="fdr")
#show results by table and plot
#phylum
#table
metatab.show(metatab=metab.sex$random,com.pooled.tab=tabsex4[tabsex4$id %in% tlm,],
tax.lev="l2",showvar="genderMale",p.cutoff.type="p", p.cutoff=1,display="table")
#plot
metadat<-metatab.show(metatab=metab.sex$random,com.pooled.tab=tabsex4[tabsex4$id %in% tlm,],
tax.lev="l2",showvar="genderMale",p.cutoff.type="p", p.cutoff=1,display="data")
meta.niceplot(metadat=metadat,sumtype="taxa",level="main",p="p",
p.adjust="p.adjust",phyla.col="rainbow",p.sig.heat="yes",
heat.forest.width.ratio =c(1.5,1), leg.key.size=0.8,
leg.text.size=10, heat.text.x.size=10, heat.text.x.angle=0,
forest.axis.text.y=8,forest.axis.text.x=10,
point.ratio = c(4,2),line.ratio = c(2,1))

metamicrobiomeR documentation built on Nov. 9, 2020, 5:06 p.m.