Dear Dr Peter and WGCNA collegues,
Thanks collegues in genetics.ucla.edu for sharing WGCNA package for co-exp analysis.
But we are confused about the page margin setting, sth like shuffling the rows and columns, mentioned in Figure2 of "Can I use the WGCNA dendrogram to display a heatmap of gene expression?" (https://support.bioconductor.org/p/67548/).
when using the function plotNetworkHeatmap, which just has several parameters (datExpr, plotGenes, useTOM, power, networkType, main) not mentioned margin setting. Thus we get the picture JUST showing the length less than 12 characters, but the other 3 more letters are outside the view. Take the gene name MSTRG.11922.5 for example, the letter "M" did not show in row label, while the last letter "5" disappear in column label in the view, all this two letters are hidden in the margins.
Regarding LOC_Os02g38440.1, "s02g38440.1" can be seen in row name list, while "LOC_Os02g384" can be seen in column name list, displaying alternative hide and show.
Despite setting par() like following, plotNetworkHeatmap output with the default parameters as it did, .
pdf(file = "signed correlations.pdf", width = 8, height = 8, paper="a4")
topList=rank(NS1$p.Weighted,ties.method="first")<=30
gene.names= names(datExpr)[topList]
op<-par(mfrow = c(1,1), mar=c(2,2,2,2), cex.axis=2)
plotNetworkHeatmap(datExpr, plotGenes = gene.names,networkType="signed", useTOM=FALSE,power=1, main="signed correlations")
par(op)
dev.off()We also try par(mfrow = c(1,1), mar=c(8,10,6,2), cex.axis=2) before or after plotNetworkHeatmap, but failed in the end, and the long gene names mentioned above displaying alternative hide and show. Nor layout(matrix(c(1,1,0,2), 2, 2, byrow = TRUE)) did succeed.
Thus we are wondering that the proper preset plotNetworkHeatmaphas has nothing to do with par() or layout().
Since we are not sure which .r file in the subdir R of WGCNA1.49 is closely related to plotNetworkHeatmap, we are wondering that the proper preset of plotNetworkHeatmap.
Therefore, we wish we can solve the problem of margin setting by deeply understanding the related settings for using plotNetworkHeatmap smoothly.
Thanks again.
