Can I used GFOLD or EdgeR to identify DEG in a time-course type experiment with no replication?

Question

[alexandra.lanot]

Dear all, I am new to Bioinformatic but I hope my question will make some sense. I need to analyse my RNASeq data: 7 samples corresponding to different developmental stages but no replication (I know that this is bad). I was thinking of using EdgeR or GFOLD. Most of the information I have seen is to compare two samples. I need to compare them as a whole as I would be interested in genes expressed in only a few of the samples.
Can I use GFOLD for comparison of more than 2 samples? Can I used EdgeR for comparison of more than 2 samples knowing I have no replication? Many thanks in advance A

[Fion]

Hi~ How’s your running results in the end? I'm also struggling in a similar situation: 4 sequencing libraries corresponding to different developmental stages, each library was constructed from total RNA extracted from pulled embryonic tissues. So the sequencing data do come from several embryos. But when conducting bioinformatic analysis, I'm not sure whether I should use NOISeq, GFOLD or EdgeR? I'm also new in bioinformatics. May I have your suggestion? Many thanks!

Answer 1

[WouterDeCoster]

Replicates are really required for differential expression analysis. Perhaps you can convince one of the tools to work for your poor experimental setup, but I would trust any data coming from that. Would it make sense in your design to compare the 3 first with the 4 last timepoints?