Hi,

For few of my plant samples with no biological replicates I am using edgeR.

If I use bcvvalue 0.1, I am getting >10k significantly differential expressed transcripts out of 47310 transcripts.

If I use bcvvalue 0.4, I am getting ~5k significantly differential expressed transcripts out of 47310 transcripts.

edgeR manual suggest that bcvvalue 0.4 for human.

I am not sure which to select? Because of more than 10K significant genes?

Can any one suggest me the right way?

Regards,

Manoharan

The best you can do is to choose a sensible BCV value based on your experience with similar plant samples where you did have replicates. In the absence of replicates, the aim is more to rank to genes rather than to obtain a rigorous measure of statistical significance.

Sorry, there's no "right" value. If you want to choose the BCV more objectively, then you would need to do replicates.