search

omnideconv / SimBu

Simulate pseudo-bulk RNAseq samples from scRNAseq expression data
http://omnideconv.org/SimBu/
GNU General Public License v3.0
12 stars 1 forks source link

Adding a TODO md file to track the potential issues - before submission to Bioconductor. #21

Closed federicomarini closed 2 years ago

federicomarini commented 2 years ago

Upon resolving, these can get a simple strikethrough (or even use checkboxes).

Hope this can help even more in grooming up before submission.

alex-d13 commented 2 years ago

I worked through all the errors, warnings and notes. BiocCheck still prints some issues, I added them to the end of the TODO file with some comments, @federicomarini could you check if I should look at some again?

One question on parallelization: I am currently using parallel::mclapply() for my multi-thread functions. Should I switch that to BiocParallel::bplapply() (https://contributions.bioconductor.org/r-code.html?q=parallel#parallel-recommendations)?

mlist commented 2 years ago

Yes, I think it will make the review process easier if we use the recommended parallelization. I guess it is not too much work to switch?

federicomarini commented 2 years ago

BiocParallel is a breeze to use, also because it automatically handles the backend to use for you. So, yes, from my side it is a totally recommended thing to do, not only because Bioc asks you to do so.

BiocParallel::bplapply() is as close to a drop in replacement as it can be. My suggestion to be sure that it is the case, do a couple of targeted unit tests before & after switching?

I'll be on the items in the TODO later 😉