Hi everyone, I have the Sn-RNA seq sequencing data for 20 sample (10x). 9 control samples were processed in 5 batches, and 10 controls+1 case was processed in one batch. Moreover, the first 9 samples were processed using chemistry 3 and second set (11 samples) were processed with chemistry v3.1. I see significant chemistry effect more than batch effect and status effect. I'm not sure how to handle this. I had no choice to do this and the only thing I could do was to include one of the cases with control set. I don't know whether it is efficient enough to remove chemistry effect. I was wondering how much I should rely on the result of removing chemistry effect that it doesn't remove the interestig biological signals? If I want to run SCTransform, should I split across samples. I believe that if I split across any variable it will be integrated across that variable accordingly. My question is that should I split into different samples or chemistry or affected status? I appreciate any help on this.
