Repository navigation
specs/048: pathogens.csv; homodimer-aware output binding (LNG_BIND_STOICH) - #110
Merged
Merged
Conversation
…joins through member/split hops The pool joins (LNG_POOL_JOIN_MEMBERS, LNG_POOL_REGEN=protein) find the MAPK1 and MAPK3 pools in RAF, which then share 246 capped intermediate nodes and are both dropped. Kept here for the shared-intermediate design. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…OICH) pathogens.csv lists pathogen-derived nodes by the drug rule with species in place of the Drug class (a leaf with a species and none human; a complex if any component is; a set if every member is), for a consumer modelling an uninfected cell. Written always, networks unchanged. LNG_BIND_STOICH=1: the output-binding signature keeps a complex's leaf multiset, as its docstring said it did. Without it a homodimer reads as its monomer and the tie is broken by sorted rank, so R-HSA-5675376 bound p-MAPK1 to the MAPK1 dimer and the p-MAPK1 dimer to MAPK1. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…gnal specs/048 arm bind: 0 predictions moved, structural checks met) Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…sv lists only entities carrying a pathogen PROTEIN; viral RNA ligands are not held Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
… adopted) Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…the query test now carries reference classes Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
…mers count via their unit); signature cache keyed on LNG_BIND_STOICH; docstring names the real RNA classes; tests for nested stoichiometry, the cache key and Polymers Checked: the Polymer change lists no new entity in any of the 92 catalog pathways (analysis/048/polymer_check.py), so the measured arms stand. Co-Authored-By: Claude Opus 5.5 (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.
For deltasignal specs/048. The arms, results and review record are in deltasignal
specs/048-experimental-levers/research.md(PR reactome/deltasignal#101).Changes
pathogens.csv, always written besidedrugs.csv, listing pathogen-derived entities:repeatedUnitcounts as a component);LNG_PATHOGEN_PROTEIN, on by default. A leaf counts only when it is a protein: an EWAS on aReferenceGeneProductorReferenceIsoform. Viral nucleic acids are not listed; in DDX58/IFIH1 they are the ligand the pathway senses.LNG_BIND_STOICH, on by default. The output-binding signature keeps a complex's leaf multiset with stoichiometry. Before this, a homodimer tied with its monomer and the sorted-id tie-break crossed the pairs at R-HSA-5675376.Measured (deltasignal, pre-registered, both against canonical)
bind: no prediction moves on either axis. It is adopted as a faithfulness fix: binding ties fall 64 → 58, and edges change in 10 pathways.pathp, and both changes combined (combo): experimental +7 (7/0, PDPK1 knockout), curator held-out +1, tuning +14. Nothing breaks, and the prediction was met exactly.Review: an adversarial review found no blocking defect. Fixed:
Polymer handling lists no new entity in any of the 92 catalog pathways. CI is green at 4a914a4.
Merge order: this PR first. Then I rebuild
currentfrom generator main and benchmark it, and deltasignal#101 merges after that, because its new defaultDS_PATHOGEN_MODE=inertneedspathogens.csvin the catalog.🤖 Generated with Claude Code