Make the alphagenome_pt null alias state what verifies it - #241
Merged
Conversation
Keeping the alias (confirmed), so the premise it rests on should be traceable rather than asserted. _CDF_ALIASES makes one backend's percentiles depend on another backend's null. That makes "the two produce the same predictions" a correctness requirement, not a convenience note -- and the old comment stated it as a bare fact, which is how it stood while being silently false for 738 splice tracks (correlation 0.20 against this very null, fixed in #240). Both comments now point at tests/test_alphagenome_backends_equivalence.py, name the tolerance it enforces (one track per output type, correlation > 0.99, peak-relative < 8%), record that agreement is ~1-5% per head as measured rather than assumed, and warn against narrowing that test's coverage again -- the earlier three-DNASE-track version passed throughout the period the claim was false. No behaviour change; comments only. Fast suite 2,142 passed / 35 skipped. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Follow-up to #240. Keeping the alias was confirmed, so the premise it rests on should be traceable.
_CDF_ALIASESmapsalphagenome_pt → alphagenome, which makes one backend's percentiles depend onanother backend's background null. That turns "the two produce the same predictions" into a correctness
requirement, not a convenience note — and the comment stated it as a bare fact, which is exactly how it
stood while being silently false for 738 splice tracks (correlation 0.20 against this very null).
Both comments now:
tests/test_alphagenome_backends_equivalence.pyas the thing that holds the claim up,throughout the period the claim was false.
Comments only, no behaviour change. Fast suite 2,142 passed / 35 skipped.