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

A question about the '--threshold-affinity' parameter #219

Closed
chenli-bioinfo opened this issue Sep 4, 2023 · 1 comment · Fixed by #225
Closed

A question about the '--threshold-affinity' parameter #219

chenli-bioinfo opened this issue Sep 4, 2023 · 1 comment · Fixed by #225

Comments

@chenli-bioinfo
Copy link

Dear author,

I was running mhcflurry using:

mhcflurry-predict-scan test.fasta --peptide-lengths 8-11 --alleles HLA-A0101 --threshold-affinity 500 --threshold-affinity-percentile 2.0 --out ./test.csv

While the percentile threshold was correctly applied, I found that affinity seemed not to be cut off:
WeChat3ee5c49739905c79444e6ef0df084400
I was therefore wondering if I was understanding the parameters correctly.

Thank you very much.

Chen

@timodonnell
Copy link
Contributor

Thanks for reporting this, Chen. This is happening because only one quantity can be filtered, so your affinity threshold is being ignored. The library API (Class1PresentationPredictor.predict_sequences) makes this clear but I agree in the commandline tool this is super confusing. I am changing this now to have different semantics for the commandline tool and to apply all the filters you specify. The fix will go into the next release.

timodonnell added a commit that referenced this issue Oct 17, 2023
Drop the --results-best flag to mhcflurry-predict-scan. Apply all the
threshold filters specified rather than just one of them. Closes #219.
The rationale for this potentially breaking change is that this it is
closer to user expectations.
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 a pull request may close this issue.

2 participants