Alignment score #80
Alignment score #80
Conversation
|
Hi all, Just wondering what the status is of this PR? What needs to be done to get it over the line? I'm happy to try and help if necessary. |
Ohhhhh I know the feeling! Thanks so much for this great library/bindings. I'm about to start a project using this library so I'll probably be harassing you a bit over the next couple of weeks 😅 |
|
If you want to take a stab at this (#71), I’m willing to help. Since I found a work around, it’s not top priority for us right now, but it seems important for the lib in general moving forward. —Rob p.s. I too know the feeling (not moving this time, but grants and conference deadlines out the wazoo this past month!) |
|
I'm going to merge this, and reconcile with with #85 and then see if I can hack out Arc and push a release |
@rob-p Dunno if you can test it, or added the proper tests? @mbhall88 Query name is passed in as AsRef<[u8]> and there is some not my favorite code in there to prevent a new allocation if the query name is already nul-allocated, and to do an allocation if needed. Also I've updated Mappings to return Arc<String>'s for query name and target name, which should decrease performance for very small jobs but increase for large jobs.... but I need to benchmark
This PR adds a function
map_with_namewhich allows resolving #75. Specifically, it passes the query name (which must be\0terminated) to the map function, which allows the internal minimap2 functionmm_mapto resolve ties in the chain score.Though the differences are subtle, this allows exactly replicating the behavior of command-line minimap2, in cases where this tie breaking can result in some alignments being marked supplementary, where they otherwise would not.