Entering edit mode
9 months ago
rohitsatyam102
▴
920
Dear Members of the community
I have been presented with a weird situation. I have RNAseq dataset for 3 genes along with respective controls. Two of them underwent knockout and one was subjected to functional knockdown. Now Knockouts looks perfectly deprived of reads when compared to control samples. But when I visualize the Knowdown samples I see read coverage in the intronic region as well not only for the gene that underwent functional knockdown but also the neighbouring genes as shown below:
Sorry - can you be a little more clear why this is a problem? A knockdown should show reduced expression, not the complete elimination of expression, so you would expect to see some coverage. The question is whether the relative (to library size) expression of the knockdown gene is lower than the control; this can't be seen directly from IGV.
The relative expression of the Knockdown compared to the control is less. The introns show depleted coverage. But that's not what worries me. My worry is why I see reads in intergenic regions and intronic regions (even after setting MAPQ > 30 filters in IGV) that are high-quality mapping reads (with MAPQ of 255) and are primary alignments!! And on top of that the neighboring genes also showed intronic reads even when they were untouched.
This may have to do with some specifics of the knockdown (for instance, if using siRNA, you could be knocking down one transcript, so you start seeing another one with retained introns) , or it could be an issue with genomic contamination. Is the intergenic% substantially higher for that batch of sequencing?
As per the image below obtained by Running RNASeQC the intronic reds aren't that much (See the knockdown samples enclosed in the Red box) (at least not more drastically different) compared to other samples that underwent gene knockout.