Open skanwal opened 5 years ago
Dear Sehrish,
Thanks a lot for your input, highly valuable! Generally, I can say that what you suggest makes perfect sense as a further development of the workflow. And parts of your ideas have been mentioned by some other colleagues here. I will get back to you shortly with my ideas/comments on what is realistic short term etc., very busy here today.
PS. You are correct about the --target
, this should refer to the targeted region of the input sample. But I have in fact not implemented this one yet, so it is currently only there as a placeholder. Will update that shortly.
regards, Sigve
Hi Sigve,
Thanks for the response and I look forward to hearing back from you. Happy to coordinate/contribute always.
Regards, Sehrish
Hi Sehrish,
Coming back to this:
Would appreciate your input on this.
regards, Sigve
Hi Sigve,
Thanks for getting back to this.
• Reference data from Hartwig is:
Also, I do appreciate the point that we need to understand what data we are going to use for presentation as it’s hard processing reference input on the fly.
• Global preview: We are hoping to begin with focussing on coding regions.
It would be definitely very useful to have the ability to switch to whole gene (including introns). But we can expand on this later.
Happy to have your feedback on this and start looking into implementation as well - if this sounds feasible/useful to you.
Regards, Sehrish
Hi Sigve
Thanks for another awesome framework. We (@umccr) are very much interested to incorporate this into our reporting.
I have looked at the github repo/code and tested it locally - it works great. We have a couple of questions/comments:
We are interested in using some of the reference data from Hartwig. Looking at the codebase, it shouldn't be a problem as the
data
directory is passed as an argument and this directory contains reference data to be used for the analysis. However, this might impact the annotations that the framework reads in from the.tsv(s)
in `cacao_utils.R for specific clinical genomic tracks?There is an optional flag
--target
, which according to my understanding refers to the targeted region in the input sample?This probably links back to point 1 i.e. feeding in one specific
bed
track (in this case) which could be joint set of various loci sources such as CIViC, CGI and OncoKb and then reading in the (optional annotations as in the code base) for this data - if this idea aligns well with your original idea of the framework?Sorry about the long commentary and thanks for your time.
Cheers, Sehrish