Change bonded host strategy to use lowest variance restraint

[IAlibay]

Fixes #1471

Developers certificate of origin

• I certify that this contribution is covered by the MIT License here and the Developer Certificate of Origin at https://developercertificate.org/.

[codecov]

Codecov Report

❌ Patch coverage is 39.79592% with 59 lines in your changes missing coverage. Please review.
✅ Project coverage is 92.37%. Comparing base (64fa847) to head (1cd1977).
⚠️ Report is 99 commits behind head on main.

Files with missing lines	Patch %	Lines
...protocols/restraint_utils/geometry/boresch/host.py	11.86%	52 Missing ⚠️
...protocols/restraints/test_geometry_boresch_host.py	82.05%	7 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1477      +/-   ##
==========================================
- Coverage   95.18%   92.37%   -2.82%     
==========================================
  Files         165      165              
  Lines       12575    12640      +65     
==========================================
- Hits        11970    11676     -294     
- Misses        605      964     +359     

Flag	Coverage Δ
fast-tests	92.37% <39.79%> (?)
slow-tests	?

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[IAlibay]

Only using a linear combination of variances here. In the past, this "worked", but probably just as well as any other approaches. Happy to change it to something else if there's a better combination we can try.

[hannahbaumann]

Thanks @IAlibay , lgtm, just the one thing. I tested this on one trajectory and it gave reasonable atoms. It does, however, take quite some time.

[IAlibay]

It does, however, take quite some time.

Do you have a ballpark timing @hannahbaumann ? i.e. is it order minute or tens of minutes or closer to hour?

This being quite expensive is somewhat expected, the number of restraints it searches through pretty much explodes with the number of potential host anchor atoms it can find. How large was the max cutoff distance for host atoms? You'd want to keep that quite small, probably no larger than 1.5 nm.

[hannahbaumann]

More like 30sec, so actually not that bad!

[IAlibay]

More like 30sec, so actually not that bad!

O yeah, I was really expecting up to 10s of minutes! I think the search space grows by ~ 10-12 potential restraints per residue (if only doing backbone atoms), so it gets to pretty large numbers quickly if we're not careful.

[github-actions]

No API break detected ✅

[hannahbaumann]

Thanks @IAlibay , lgtm!