ctb
commented
3 years ago ok this was surprisingly easy, once I remembered that all of the genbank matches are saved into outputs/genbank/{sample}.x.genbank.matches.sig. And hey, they even have decent names!
Then a quick sourmash compare outputs/genbank/SRR1976948.x.genbank.matches.sig -o cmp plus sourmash plot cmp later, and you get:
Idle musings --
- thresholding to create clusters would be good, I did some experimenting with this over in https://github.com/ctb/2017-sourmash-cluster I think
- within that thresholding, would be nice if sourmash plot supported eliminating all singleton clusters in the output...
- could probably do an automated thing to replace the labels for sourmash plot with the nicely formatted species names that we're retrieving from genbank for genome-grist
it should be fairly easy to cluster gather matches into "species groupings", for e.g. doing combined sgc queries.