fold-change and FDR p-value threshold for miRNA (small RNA-Seq) data

Question

[Edalat]

HI, I am analyzing miRNA (small RNA analysis) data and I would like to establish a fold-change and p-value threshold which is biologically significant.with Fold change ≥ 1.5 and FDR P-VALUE ≤ .05, I lose most of miRNA.

Answer 1

[kanwarjag]

Fold change may not be best as some microrNAs will have zero values in one case while very large value in another case. you can still calculate the FC by transforming values, but I think p value may be best, you can use DEseq, EdgeR or even simple t test.

[Edalat]

thank you for attention. with FDR P-VALUE ≤ .05 of t-test (without fold change) , I lose most of miRNA too, I think to use bigger p value?

[kanwarjag]

Then I will see how many replicates you have. It looks like you have few replicates and which do not show a good correlation with in a group

[Edalat]

excuse me which is better for filtering miRNA ,FDR P-value or P-value?

[zizigolu]

I read many things about being cautious with p-value then I prefer FRD