Closed smoe closed 2 years ago
@borauyar Corrected. Thanks for spotting that. Have rebased to master but admittedly I still did not run this. Do you have an easy check on your end?
@smoe thank you! I will check this and other recent changes altogether. I think this should run through. If something fails, I will fix it.
Your reports are the first encounter for our collaborators with their data. Our experimental setup is a bit complicated and so I came up with several analyses. These all have their own name but I wished for a description in the report that explains what the respective comparison is about.
I have not run this locally, yet (snakemake is still busy) but thought I seek this early opportunity to ask if you like this idea, so I will improve on it.
Since the reports are invoked both on salmon and hisat2, I think it would be nice to see explained in the introduction what the particular value of each quantification is. Maybe I would even want to go as far as to provide an automatically generated methods section at the end of the report.
Would that be something you would lke to see as a PR (maybe an evolution of this one)?