search

maickrau / GraphAligner

MIT License
256 stars 30 forks source link

considerations for large consensus genome graphs #30

Closed ekg closed 3 years ago

ekg commented 3 years ago

The consensus graphs from pggb are designed to be used by alignment tools like GraphAligner. They have large nodes, and simple topology. So we're testing GraphAligner but I had a question about the preparation of the graph. Is it beneficial to sort the graph topologically, so as to improve seeding?

maickrau commented 3 years ago

The node ordering or names won't matter. GraphAligner internally finds "mostly acyclic" subgraphs based on the graph topology and then seeds along those.

ekg commented 3 years ago

Thanks, it's great to know it's doing that.