Closed ckeeling closed 4 years ago
Hi @ckeeling,
Thank you for asking! The gene coexpression relationships are connected by the similarity of expression profiles (expression levels of different time points) between any pair of genes, so you can use either RPKM or VST read counts as expression level to do the TO-GCN analysis.
petitming
Did anyone try with VST read counts yet?
It was strange that I got the Segmentation fault (core dumped)
in Cutoff step only when my inputs were VST read counts.
All_genes_vst.tsv.zip
Intersection_TF_vst.tsv.zip
However, when I took the normalized read counts generated from DESeq2 as inputs, everything worked just fine. Intersection_TF.tsv.zip All_genes.tsv.zip
@petitmingchang @ckeeling Have you ever encountered this issue?
UPDATE: Cutoff using TPM values worked fine, too.
Hello @petitmingchang,
I use DESeq2 for RNAseq analyses, which doesn't generate RPKMs. In fact, it is oblivious to transcript length in its count-based analysis. Will your analysis be applicable if Variance stabilizing transformed counts are used instead of RPKMs?
http://bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#variance-stabilizing-transformation
Thanks!