Fix very high false negative rate for insertion SVs with LRG

vgteam / long-read-giraffe-experiments

Experimental materials for testing the long-read mapping abilities of vg giraffe

Other

4 stars 0 forks source link

Fix very high false negative rate for insertion SVs with LRG #55

Open adamnovak opened 1 week ago

adamnovak commented 1 week ago

Calling with Long Read Giraffe gets @jmonlong a very high false negative rate for insertion SVs. We need to figure out where they went.

See https://ucsc-gi.slack.com/archives/CJ2EHEH1A/p1729633305487609

xchang1 commented 1 week ago

@jmonlong I think I fixed this here The problem was that when we tried to align between two nodes against the graph, we pulled out a subgraph that excluded the insertion node because the distance to go between the seeds with the insertion would be longer than the length of the read subsequence we wanted to align. I changed the distance limit we can walk through to find the subgraph to include the length of the longest gap we allow