I have total 6 samples (3 controls and 3 treated) and have a total of 576149 genes to be compared between control vs treatment. However, I found a sub set of 320336 genes for which five samples have 0 count and randomly any one of the sample has count greater than zero. Keeping these 320336 genes in my data set for is not giving any adjusted p-value significant (even I increase the cut off threshold 0.1) after DESeq2 run; however, removing those genes helped to identify differentially expressed genes with adjusted p-value <0.05 and we are able to capture relevant biology. By the way, independentFiltering option was kept TRUE by default. Do you think that I can remove those genes (counts for all 5 samples=0 and randomly one sample>0) before I run DEseq2.
Thank you in advance.