Complex II: enables → contributes_to

A relation fix for succinate dehydrogenase subunits

AI Gene Review · projects/SDH_GP2TERM_CONTRIBUTES_TO · in progress · 2026-10-05

Bottom line

  • No single subunit of succinate dehydrogenase performs GO:0008177; GO agreed (go-annotation#6414) that subunits get contributes_to, not enables.
  • 7 reviews carry GOA rows for GO:0008177. SDHC was already structured; SDHB/SDHD now use MODIFY + paired NEW contributes_to rows.
  • Still open in #582: SDHA's GO:0000104 half-reaction decision, 9POAL NCGR_LOCUS67308, and PSEPK sdhA/sdhB.

The mechanism behind the rule

Original human and 9POAL audit

The SDHA nuance

  • SDHA's FAD site can oxidise succinate with an artificial acceptor: the half-reaction.
  • That matches the parent GO:0000104 succinate dehydrogenase activity, so enables there is defensible for SDHA alone.
  • The quinone-coupled GO:0008177 still needs the whole complex: contributes_to for all four subunits.
  • The current SDHA review MODIFYs three GO:0008177 rows to GO:0000104 and keeps one as contributes_to: this needs an explicit decision.

Status and next steps

  • ✅ Tier 1 (PR #3223): SDHB (3 rows) and SDHD (1 row). The GOA rows keep GOA's own enables qualifier (not overridden); each enables row is MODIFY and a paired NEW contributes_to row (IDA) records the intent. Re-fetch once GOA reflects #6414.
  • ⬜ Tier 2: decide SDHA GO:0000104 vs GO:0008177; decide 9POAL NCGR_LOCUS67308 (MODIFY → GO:0009055 or keep with qualifier).
  • ⬜ PSEPK sdhA / sdhB reviews carry GO:0008177 with enables; apply the #6414 pattern.
  • Order: MODIFY + paired NEW row → wait for GOA → just fetch-gene → just validate.

Read more: projects/SDH_GP2TERM_CONTRIBUTES_TO.md