Combat-seq, Pseudoaligners and DESeq2

[pintoa1-mskcc]

Hi

I have been working with combat-seq for a little while, where I quantify with kallisto, import with tximport, load to deseq2 to get "raw" counts, then run combat-seq on that data.

I noticed that when i use combat_seq, I need to load my data into deseq2 as a matrix, which removes all the special normalization deseq2 will do with tximported data.

Would this be the proper way to analyze pseudo counts? I feel I am losing a lot of the transcript/size factor normalization that is expected with pseudoaligners when I feed the data into combat-seq this way