Closed adamnovak closed 2 months ago
Only ~60% of nanopore reads seem to be mapped after surjection. It looks like we are hitting a limit on the graph size for the subgraph we surject onto and bailing out. We should probably just limit alignment size instead like Jordan suggests.
I think I fixed this in vg a6dfc43d, by just scaling the existing threshold by the read length.
Only ~60% of nanopore reads seem to be mapped after surjection. It looks like we are hitting a limit on the graph size for the subgraph we surject onto and bailing out. We should probably just limit alignment size instead like Jordan suggests.