Hi, I have used voom from the "Differential count" package and it worked smoothly. I would like to get the matrix of the adjusted count values, which is generated by the voom package. These would be good for e.g. PCA etc. How can I extract these adjusted/weighted count values from Galaxy? Thank you so much for your consideration.
I assume you're referring to https://email@example.com/repos/fubar/differential_count_models ? My limited understanding of http://genomebiology.com/2014/15/2/R29 is that voom generates 'precision weights' for each individual normalised count, not "adjusted" counts. I'm not sure those precision weights would be useful for MDS or other downstream uses other than voom itself which the tool already offers but am interested to hear of other uses if you know of any?
That said, the weights could be another optional output if voom is selected if enough people want them - may I ask anyone who would like those weights to be an optional output to add a comment saying so here please? It won't take much urging because it's a trivial change but I'm not sure how useful they really will be and I'd rather not add yet another option to what is a fairly complex tool interface unless there are at least a few more users who want them enough to speak up.
The good news is that in the meantime, because the actual generated R code is downloadable from the HTML output of any successful run, if you snarf that R script, you can add a single write.table(...) line and rerun it on the command line using Rscript to get what you need until the tool is updated.
Thanks for using the tool!
Hi Fubar, I am using the tool through server GVL/QLD. I really appreciate all the great work by you and any team to provide access to the tools.- I believe that currently the voom subtool of the "Differential_count model" tool takes the raw count data (which one has access to, of course), and delivers a result table on, say modified, adjusted, normalized, you name it, values. It would be a helpful add-on, and I think necessary add-on, to get the actual normalized values, so one would not have to leave the platform. It is an excellent idea, though, to provide the R code. Thanks again. Cheers.