Falcon deep research succeeded on the second attempt (IFT81-deep-research-falcon.md; first run failed with falcon 429/timeout and an unavailable perplexity-lite fallback; rerun with --timeout 2400). It concurs with the tubulin-module model [file:human/IFT81/IFT81-deep-research-falcon.md "Depleting IFT81 in human RPE-1 cells strongly reduced primary-cilium formation; siRNA-resistant wild-type IFT81 restored it, whereas a mutant disrupting the full tubulin-binding patch failed to rescue."] and adds Boegholm et al. 2023 (not cached): the IFT81-IFT74 coiled coil stimulates RabL2 GTP hydrolysis [file:human/IFT81/IFT81-deep-research-falcon.md "The activity explains RabL2 inactivation and release shortly after IFT initiation."]. This supports MODIFY of RABL2B 'protein binding' rows to small GTPase binding; a GTPase activator annotation is raised as a question rather than asserted (data mainly from Chlamydomonas proteins). It also cites Perrault 2015 (NPHP with polydactyly, IFT81 variants).
HPA v25: Primary cilium (Approved), Basal body (Approved), Centrosome (Approved); main location Cytosol. The cytosolic main call reflects the large unassembled/cytoplasmic pool; ciliary and basal body calls fit the IFT-B role. No GO_REF:0000052 rows. Module role "IFT-B core; tubulin binding" and role_description "IFT81-IFT74 bind tubulin cargo" are fully consistent with core_functions (MF tubulin binding; intraciliary transport involved in cilium assembly; anterograde IFT; IFT-B complex). IFT81 (with IFT74) is the one IFT-B subunit pair for which the module's "IFT cargo adaptor" function is directly demonstrated.
Every IFT-B subunit annotated from PMID:26980730 (Katoh et al. 2016, an IFT-B architecture paper) carries an IDA to GO:0030991 intraciliary transport particle A. This contradicts the paper's own content (which supports the GO:0030992 IFT-B IPI rows from the same paper) and the established biochemistry. Marked REMOVE as a genuinely contradicted claim (likely batch term-selection error), not as second-guessing assay data.