This is a test version of Biostars. For the public version, visit https://www.biostars.org.
What analysis suitable to identify similarly expressed genes between two samples (treated vs untreated). Contrary to DGE analysis,

Hello everyone!

I'm comparing the expression of sets of miRNA data in order to identify what are the miRNAs that have significant similarity in their expression in two different samples.

I have conducted DGEs using NOISeq, and was succeeded in identifying significant differential expressed genes. By filtering out those that are not significantly different, I don't think that it is correct to assume that they are similarly expressed. Hence, i would like to know what is the most suitable analysis for this case? It would be helpful too if you could include the name of package for such analysis.

I hope someone could help me out with this.

Thanks in advance.

gene-expression rna noiseq dge
• 1,186 views • link
updated by ATpoint ATpoint • written by alifafiq1 • alifafiq1

1 answer

DESeq2 has a "lessAbs" test which gioves you a pvalue that tests whether a fold change is below a user-defined threshold, see https://bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#tests-of-log2-fold-change-above-or-below-a-threshold

There is something that you can do roughly in limma for this (https://support.bioconductor.org/p/64300/) but to my knowledge this has not been formally implemented, so might or might not work well, I cannot tell.

• 0 views • link
written by ATpoint ATpoint

Thank you @AT point. I was planning to use DESeq2 and use the alternate hypothesis (lessAbs) too considering that is the only package that could do so via DGE.

However, some of my samples have no replicates and i was planning to conduct analysis (including with those without replicates) just to explore their results since I know that analyzing no replicates samples won't yield strong and publishable statistical data. And also, DESeq2 doesn't support no replicates data. Hence I wonder what could be the alternative to this.

• 0 views • link
written by alifafiq1 • alifafiq1

Can you post your setup, so how many groups and their replication number? I can tell you by experience that you need even more replication to robustly identify similar rather than differential genes, so your analysis might not be feasible.

• 0 views • link
written by ATpoint ATpoint

Log in to answer this question.

More posts like this