Entering edit mode
2.9 years ago
felipead66
▴
120
In search for possible 3'bias I have run gene_body_coverage.py and come up with a plot that assimilates the magenta plot in this example:
Does this mean there is 3'bias? What does the "shoulder" around 70 suggest?
For me, the magenta line is fine, there is no 3' bias at all. If by "shoulder", you mean the small bump around 70%, I don't think you should read too much into it. There is always some stochastic variation. Perhaps it can be explained by a small fragmentation bias but it is difficult to prove.
Thank you for your reply. Is there an explanation why a group of samples follow the same line? In this case I am talking about magenta, blue, purple etc. In my case this is the trend as well. They comprise of the same cells with a different treatment. Is it normal to have the same trend in all plots?
This is RNA-seq right ? In RNA-seq, highly expressed genes are over-represented in those kind of metaplots. Small sequence bias (GC content, etc...) in those genes affect their coverage, which explain the small variations, since those genes contribute for most of the average profile. Because in most conditions, the same genes are highly expressed, you have the same trend across conditions.