This is not a banal discussion. I am facing some problems with the analysis of DE genes in mouse.
Most methods of analysis of DE genes must face two considerations or challenges. The first needs to take into consideration the existence and the different isoform expression of the genes, and the second the variability obtained across biological replicates. These two needs to be accounted together. In the first case, the collapsing or summarizing of the raw isoforms counts to the genes notably affect the output of my analysis
My DE analysis are facing a differential output depending upon the program or package I am using. And I need to take into consideration whether I should generate pseudo-aligned counts or mapped BAM files to this thread.
Most packages face this analysis by considering one of these challenges, but not the other. The pletora of R packages available and/or standalone programs to this end, is high, and I need you to participate in this discussion to help me and others to make the best choice for this type of analysis
To mention, I started to use Kallisto and DESeq2 to see that after importing the data with the tximport package using the txOut = FALSE or TRUE gives two completely different outputs, one being that not DE genes are expressed, and the other giving rational but inaccurate DE expression
¿What do you recommend?