Entering edit mode
5.7 years ago
nazaninhoseinkhan
▴
520
Dear all,
I have a large list of DEGs (around 400). However when I tried to do pathway enrichment analysis using DAVID, PANTHER or Enrichr, based on FDR only 1 pathway was detected.
Can u tell me the probable reason for this? Is there anything that I did not consider?
Looking forward to your comments
Nazanin Hosseinkhan
Please give information on the experimental setup, the number of samples, treatment and the tool that was used for differential analysis.
52 tumor samples vs 3 normal tissues belong to TCGA project were compared using DESeq2 program.
The comparison is imbalanced - which TCGA dataset is it? Are you using FPKM data? FPKM data is not suitable for cross-sample differential expression analysis. From where did you obtain the data?
Hi Kevin,
I imported a list of deregulated genes( 30 down vs. 48 up-regulated) resulted from DESEq2 into Enrichr. However, no FDR significant pathways were detected. I checked some of my genes in KEGG and strangely they were detected in the same pathways. I could not understand the meaning of FDR in this context, i.e. pathway enrichment. Could you please explain to me why no pathway has significant FDR in Enrichr.
Thank you in advance Nazanin
I see. 30 + 48 genes is not that much for gene enrichment or pathway analysis, which is probably why you are not obtaining any significant adjusted P values. With that small number of genes, have you tried to just do a literature search to infer (yourself) in which pathways these genes belong?
Enrichr identified the same pathways as KEGG. I manually searched some of my genes by KEGG. So is it correct if I ignore the FDRs reported by Enrichr?
You can choose to ignore them, if you wish, but you should report some enrichment score and/or unadjusted P value. If you publish the data, you may be questioned by a reviewer / asked for clarification on why you did not use the adjusted P values.
I still believe that a full day of literature searching involving your gene of interest would be more productive. You only have 78.
Sure.
I will literature searching them. If I understood correctly, the KEGG has not been updated since 2016 and if I do literature searching I might found more pathways than KEGG.
Thank you again for your prompt response