## EMC6
- **UniProt:** Q9BV81 (TMEM93) · **batch:** proteostasis-batch-2026-06-11 · **review status:** COMPLETE (thorough; all annotations including autophagy block adjudicated)
- **PN placement:** Row 1 `ER proteostasis | Protein transport | Transmembrane protein import | EMC complex component`; Row 2 `Autophagy-Lysosome Pathway | Autophagophore initiation and elongation | Class 3 PI3K complex 1, direct | Localization of class 3 PI3K complex 1` (branches ER + ALP). **PN-node mapping:** row1 type→GO:0072546 EMC complex (already_in_goa_exact), group→GO:0044743, class→GO:0015031; row2 entirely context_only/too_broad_to_propagate (GO:0035032 PI3K III; GO:0016236 macroautophagy) — projects NOTHING.
- **Consistency:** Consistent, and the two-branch placement is well-handled on both sides. EMC6 is the catalytic-core subunit pairing with EMC3 in the vestibule (D27/T31 mutagenesis separates insertion from assembly → GO:0032977 core). Separately, the single 2013 study (PMID:23182941: RAB5A/BECN1 interaction, DFCP1/omegasome colocalization, autophagosome-formation defect) underlies GOA's GO:0000045 (autophagosome assembly, IMP+IBA) and GO:1903349 (omegasome membrane, IDA) — all KEEP_AS_NON_CORE in the review, matching PN's judgment that the autophagy role is context, likely indirect via client biogenesis. No contradictions.
- **PN story / NEW pressure:** Row 2 is the only place PN asserts a role beyond the EMC insertase story (BECN1-mediated recruitment of PI3K complex to the autophagophore nucleation site). This is partially captured in GOA (autophagosome assembly, omegasome membrane) and in the review (BECN1/RAB5A protein-binding, omegasome). PN correctly does NOT claim EMC6 is a PI3K-complex member (GO:0035032 marked too_broad/context_only). No defensible NEW GO term — the existing autophagosome-assembly term covers the asserted role and is already present. Conclusion: already captured (and appropriately non-core).
- **Mapping strategy:** EMC6 does not change either node. Row1 type→GO:0072546 exact/correct; projected group/class (GO:0044743, GO:0015031) broader than review's specific insertion terms (broader-ancestor pattern, cf. TOMM20/HSPA8/RAB7A). Row2 correctly projects nothing — EMC6 is a BECN1 interactor/regulator, not a class-III PI3K complex component, so withholding GO:0035032 membership is the right call (avoids a false complex-membership assertion).
- **Evidence alignment:** Good overlap on EMC papers (PMID:22119785, PMID:29242231, PMID:32439656, PMID:34918864). Row2's PN reference is a non-PMID review ("Membrane Trafficking in Autophagy - ScienceDirect"); the review instead anchors the autophagy role to the primary PMID:23182941, a stronger citation. Slight divergence: PN row-2 evidence is a secondary review whereas the gene review uses the primary experimental paper.
- **Verdict:** Consistent across both branches; autophagy/PI3K role correctly captured as non-core and correctly NOT projected as complex membership; PN adds no NEW pressure; row1 projected group/class terms broader than review.