Entering edit mode
I have multiple datasets that I applied limma's differential expression analysis. To present the genes, I found that the MetaVolcanoR
package is suitable.
But, I don't understand how can I tune the parameters of the plot. How do I tell it what Mean Fold change (the x-axis) is significant? like how do I determine the threshold?
And the same for the P value ( y-axis), I see there is a threshold around 0.5 in this plot, how do I know for sure what threshold it's using?
This is my code: (the deg
data frames are the results, containing all the genes for each data with the p-value and log fold change):
totalDEG = list(table4 = deg4, table6 = deg6,
table7 = deg7,table8 = deg8 ,
table15 = deg15, table21 = deg21)
totalDEG = map(totalDEG, ~ .x %>% rownames_to_column("symbol") %>% `row.names<-`(.$symbol))
meta_degs_comb <- combining_mv(diffexp=totalDEG,
pcriteria='adj.P.Val',
foldchangecol='LogFC',
genenamecol='symbol',
geneidcol=NULL,
metafc='Mean',
metathr=0.03,
collaps = TRUE,
jobname="MetaVolcano",
outputfolder=".",
draw='HTML')
plot(meta_degs_comb@MetaVolcano )+ ylab('-log10.Pval')
Removed
EnhancedVolcano
tag.