Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

scoring parameter updates for broader diversity and alignment mode optimizations #247

Merged
merged 13 commits into from
Jun 12, 2024

Conversation

ekg
Copy link
Collaborator

@ekg ekg commented Jun 2, 2024

No description provided.

@ekg
Copy link
Collaborator Author

ekg commented Jun 2, 2024

The gist of this is that these are the easiest scoring parameters that actually utilize the affine gap features of WFA, which is used inside of wflign, of wflign itself, and of the biWFA-based patching. The parameters used for the patching are borrowed from minimap2's defaults, and also appear to work well.

These are:

    --wfa-params=[mismatch,gap1,ext1] score parameters for the wfa alignment (affine); match
                                      score is fixed at 0 [default: 2,3,1]
    --wfa-patching-params=[mismatch,gap1,ext1,gap2,ext2]
                                      score parameters for the wfa patching alignment (convex);
                                      match score is fixed at 0 [default: 3,4,2,24,1]
    --wflign-params=[mismatch,gap1,ext1]
                                      score parameters for the wflign alignment (affine); match
                                      score is fixed at 0 [default: 2,3,1]

@ekg
Copy link
Collaborator Author

ekg commented Jun 2, 2024

The problem that's being resolved here is that we were unable to align things at divergence of 20-30% and the alignments we were getting would often be extremely fragmentary, only including the very high identity matches and missing a lot of the more rapidly evolving intronic sequences. With this adjustment we at least align everywhere, however more testing is probably required to see if that is tending to be correct, or if we're forcing alignments through regions that don't necessarily support them.

@ekg ekg changed the title fix easy scoring parameters that work well across diverse contexts scoring parameter updates for broader diversity and alignment mode optimizations Jun 2, 2024
@ekg ekg merged commit 9ff0452 into main Jun 12, 2024
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant