Hi <@UE6D8290F> it is a relief to hear that you ar...
# references
h
Hi @Maxime Garcia it is a relief to hear that you are working on this! I wonder how do you plan to get the input files (fasta>f) into the S3 bucket in order to run this workflow? how about a module for genomepy to download any assembly name from a given provider and then run the remaining tasks of this workflow
m
Currently for this v1, is to have an asset file per genome similarly to https://github.com/nf-core/references/blob/dev/assets/genomes/homo_sapiens/ucsc/hg19.yml Minimal requirement is just genome and fasta For the moment my POC encompass sarek refs which does not need gtf, but I'll work on rnaseq soon enough, and I'll check out more ideas, genomepy looks interesting, and I'll definitively look into it, thanks for the tips. My whole idea for this v1, is just to update what we currently have in igenomes. For v2, my views are quite bigger, and @Jonathan Manning, @Edmund and other smart people are already aligning some ideas
e
Genomepy is slick, but I don't see the point in having a module for it when we can just directly link to the files ourselves. @Husen Umer we can use it to find the references we want to include though. Feel free to open any issues/PRs for references you want to see.
party this 1
h
sounds reasonable for v1, i don't have a specific species in mind but I am thinking of more flexibility especially when it comes to rnaseq in a way that the user just needs to specify a species and all files become available (could be downloaded locally to avoid filling up the S3 bucket)