I'm a bit puzzled since my last alignement with STAR resulted in high % of unmapped reads. After investigation it appears that a large portion of the reads mapped to Mitochondrial DNA. I guess it could indicate a poor cell viability before library preparation right?
My question is ; how come that in several R packages use for analysing quality of the cells in scRNAseq, we have to look at the proportion of mitochondrial RNA, whereas the aligner (STAR) already got rid of the reads corresponding to mitochondrial RNA because they didn't match to Grch38?
It seems a bit contradictory to me.