Multimodal Genotype-Phenotype Analysis in SMARCB1-Associated Developmental Disorders.

Pathogenic variants in N-terminal (winged-helix/SNF5) and C-terminal (αC-helix) regions of the BAF chromatin remodeling complex

31 individuals with pathogenic SMARCB1 variants were analyzed using a multimodal approach integrating 3D protein modeling, GestaltMatcher, and machine learning. Variants cluster in two regions: N-terminal and C-terminal (αC-helix). C-terminal variants associate with more severe speech delay, microcephaly, and cleft palate, with stronger facial gestalt similarity. An XGBoost model achieves 96.7% accuracy in predicting variant location from phenotype alone.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

Integrating machine learning into SMARCB1 genotype-phenotype correlation is methodologically interesting and clinically relevant: predicting variant region from facial features and clinical traits could guide targeted sequencing. The 96.7% accuracy is impressive but needs validation in independent cohorts.

Clinical impact: 2/3 · Evidence strength: 2/3 · Novelty: 2/2 · Sample size: 0/1 · Publication status: 1/1 → Total: 7/10