Hi All, I have received read counts per gene (48K genes) per sample after RNASeq data processing. Now I need to create heatmaps for these samples. Here is what I did: I had two types of stem cells (iPSC and ES cell) which were differentiated to three lineages in parallel (iPSC-0, iPSC-1, iPSC-2; ES-0, ES-1 and ES-2). I have 2 questions: 1. I want to analyze the read counts by DESeq2 but to my knowledge you can only do paired comparison against a control sample (not sure though). What if I want to find DEGs across several differentiations’ states? 2. I also need to create heatmaps by pheatmap after doing rlog-transformation of dds and I want to show more than 2 differentiation state in one heatmap. Since the rlog calculation is independent of DESeq2 results, I through it would be OK to create a matrix containing the read counts from ALL the samples and then load it in R for rlog transformation. Here is the code I use for rlog:
rlog transformation of the read counts from all samples:
AllCounts_Matrix <- read.table(“AllCounts.txt”, header = TRUE)
rld = rlogTransformation(dds)
mat <- assay(rld)
write.csv(mat, "rlog_AllSamples.csv")
Load rlog data for heatmap:
rlogData <- read.table(“rlog_AllSamples.txt”, header = TRUE)
pheatmap(rlogData, scale = "row", show_rownames = FALSE, cluster_cols = FALSE)
Since I am new to R and RNAseq data analysis, I am looking for expert opinion on this. Many Thanks!
