how to deal with strong sample heterogenous effects in RNAseq

Question

[dew]

Dear all,

How to deal with the downstream digging when the RNAseq samples are so far from each other by cluster and few significant DEGs after the filter? The few significant DEGs seem not solid(enough) to explain the biological process.

Thanks a lot!

Answer 1

[i.sudbery]

You can try to correct for experimental factors, use paired/repeated measures experimental designs, find latent factors that correlate with the biological condition of interest etc. But in the end if the signal is weaker than the noise, you can't magically make it stronger.

[dew]

Thanks a lot!