Overview
Both providers frame Autosomal Dominant Cerebellar Ataxia Type III as Harding's (1982) clinicophenotypic class of "pure" autosomal dominant cerebellar ataxias — a genetically heterogeneous group of SCA subtypes whose shared endpoint is selective Purkinje cell degeneration and progressive cerebellar ataxia with minimal extracerebellar involvement. Falcon (pathophysiology-focused, Fujioka 2013-anchored) organizes the report around each SCA subtype and its molecular chain; OpenScientist is broader and more recent, adding iPSC and mouse-model mechanistic detail, epidemiology, diagnostics, and treatment evidence.
Agreement
The reports concur on the core definition, autosomal dominant inheritance, and the four established causal genes (SPTBN2/SCA5, CACNA1A/SCA6, TTBK2/SCA11, BEAN1-TK2/SCA31). They agree on subtype mechanisms: SCA6 CAG polyglutamine expansion in the CACNA1A P/Q-type calcium channel; SCA31 non-coding (TGGAA)n pentanucleotide-insertion RNA toxicity with a strong Japanese founder effect; SCA5 beta-III spectrin destabilization of EAAT4; and SCA11 TTBK2 haploinsufficiency disrupting ciliogenesis. Both converge on a unifying model in which divergent molecular triggers funnel into Purkinje cell dysfunction/degeneration, cerebellar cortical atrophy, and slowly progressive ataxia, and both agree there is no disease-modifying therapy.
Divergence
The providers differ on the roster of subtypes: Falcon includes the mapped-only loci SCA26 and SCA30 and omits SCA14, whereas OpenScientist adds SCA14 (PRKCG) and drops SCA26/SCA30. On SCA6 channel biology they diverge in emphasis — Falcon cites an older hypothesis of reduced Ca2+ influx leading to neuronal death, while OpenScientist frames excessive Ca2+ entry and excitotoxicity, and only OpenScientist elaborates the bicistronic alpha-1ACT transcription factor and the non-ubiquitinated cytoplasmic Purkinje-cell aggregate pathology. OpenScientist contributes substantial unique content: patient iPSC Purkinje-cell findings (thyroid-hormone-dependent degeneration rescued by TRH/riluzole), pre-motor cognitive deficits in SCA6 mouse models, quantitative rehabilitation and aerobic-training trial data, and detailed diagnostic/epidemiology tables. The divergences are coverage/recency and framing, not direct factual contradictions.
Integration
The concordant subtype mechanisms (SCA5/SPTBN2, SCA6/CACNA1A, SCA11/TTBK2, SCA31/BEAN1-TK2) and the convergent Purkinje-cell-degeneration pathway map onto existing pathophysiology anchors in kb/disorders/Autosomal_Dominant_Cerebellar_Ataxia_Type_III.yaml and are suitable for integration once PMIDs/DOIs are fetch-verified. The pure cerebellar phenotype (gait/limb ataxia, dysarthria, nystagmus, cerebellar atrophy) and autosomal dominant inheritance are supported by both providers.
Not integrated (leads)
Retained as leads rather than promoted: the SCA6 bicistronic alpha-1ACT and cytoplasmic-aggregate detail, the iPSC thyroid-hormone/TRH-riluzole rescue, the pre-motor cognitive/spatial-navigation phenotype from mouse models, and the specific rehabilitation/aerobic-training and symptomatic-drug trial results. These are single-provider (mostly OpenScientist) and depend on model-system or small-cohort evidence that needs primary-source verification before curation.
Cross-provider synthesis comparing the falcon and openscientist deep-research reports for ADCA type III. best_matching_text values are verbatim excerpts from the cited report files (whitespace-normalized). No direct contradictions were found; divergence is subtype-roster composition, SCA6 calcium-flux framing, and coverage/recency (OpenScientist adds iPSC/mouse-model mechanism, epidemiology, and rehabilitation-trial detail). Header MONDO is taken from the dismech disorder entry (MONDO:0019793); OpenScientist alternatively reported MONDO:0016560. Literature evidence snippets are intentionally left for the main curation pipeline pending fetch-reference verification.