Improve read mapping rules

rotary-genomics / rotary

Assembly/annotation workflow for Nanopore-based microbial genome data containing circular DNA elements

BSD 3-Clause "New" or "Revised" License

2 stars 1 forks source link

Improve read mapping rules #125

Closed LeeBergstrand closed 5 months ago

LeeBergstrand commented 5 months ago

pull read mapping code out of rule polish_polypolish and move it to a new rule.
utilize the multiext() function for removing read mapping index files (improved readability and caching).
move from bwa to bwa-mem2, the latest bwa mapping algorithm (faster and much smaller index on disk)
trick polca.sh into using bwa-mem2 by symlinking bwa-mem2 to bwa.
add config flag to keep read mapping bam files used final read coverage statistics.

LeeBergstrand commented 5 months ago

@jmtsuji I just made some significant changes to the read mapping rules. I am looking forward to your review.

LeeBergstrand commented 5 months ago

OK, let's switch over to pypolca! (I also was also thinking it would be nice to eventually switch away from using POLCA.) I will leave it up to you whether we should make the switch in this PR versus wait until a future PR.

@jmtsuji Can you do the swap out to pypolca in this PR?

jmtsuji commented 5 months ago

OK, let's switch over to pypolca! (I also was also thinking it would be nice to eventually switch away from using POLCA.) I will leave it up to you whether we should make the switch in this PR versus wait until a future PR.

@jmtsuji Can you do the swap out to pypolca in this PR?

As per our DM discussion, let's keep the bwa-mem2 replacement code and continue to use POLCA for the time being. We can look into swapping to pypolca in future -- I'll create a separate issue for it. I imagine that bwa-mem2 will probably make little to no impact on the POLCA results, but I'll at least benchmark this on one example genome for now, when I am able to do so... I'll hold off on merging this PR until I have a change to run this single genome test. As we discussed, I'll take over the final merge of this PR.

LeeBergstrand commented 5 months ago

OK, let's switch over to pypolca! (I also was also thinking it would be nice to eventually switch away from using POLCA.) I will leave it up to you whether we should make the switch in this PR versus wait until a future PR.

@jmtsuji Can you do the swap out to pypolca in this PR?

As per our DM discussion, let's keep the bwa-mem2 replacement code and continue to use POLCA for the time being. We can look into swapping to pypolca in future -- I'll create a separate issue for it. I imagine that bwa-mem2 will probably make little to no impact on the POLCA results, but I'll at least benchmark this on one example genome for now, when I am able to do so... I'll hold off on merging this PR until I have a change to run this single genome test. As we discussed, I'll take over the final merge of this PR.

Sounds good!

LeeBergstrand commented 5 months ago

@jmtsuji Are we ready to merge this yet or are you still working on it?

jmtsuji commented 5 months ago

@LeeBergstrand Still working on it -- I'd like to do a quick single-genome benchmark of BWA vs BWA-MEM2 with POLCA before merging, if that's okay.

jmtsuji commented 5 months ago

@LeeBergstrand I compared POLCA with BWA vs. BWA-MEM2 using the "Ca. Chx. allophototropha" genome. POLCA was run after polishing of the genome by Polypolish. The resulting polished genome files were identical whether BWA or BWA-MEM2 was used. Because of this, I will go ahead and merge this PR.