Purpose: To consolidate and refine RB1 disease-causing variants in retinoblastoma (RB) and to develop a visual, integrative tool for variant exploration at the gene and protein level. Methods: RB1 variants associated with RB were compiled from the Leiden Open Variation Database, 13 published cohort studies between 2014 and 2025, and the Catalogue of Somatic Mutations in Cancer. Case information was recovered where available, and variants were reannotated using Human Genome Variation Society (HGVS) nomenclature on transcript NM₀00321. 3 and mapped to Genome Reference Consortium Human Build 38 (GRCh38). PANORAMA was built in Python using Dash and Plotly. Results: After filtering for variants with unique identifiers and interpretable descriptions, 3004 variants (1943 germline, 797 somatic, and 264 variants of unknown origin) underwent HGVS standardization, GRCh38 mapping, and statistical analysis. Truncating germline variants accounted for more than 80% of variants, whereas missense variants comprised 9. 2%. Germline variant density was highest in exons 4 and 15 and lowest in exons 25 to 27. Nonsense variants predominated in the N-terminal, Pocket A, and C-terminal regions, whereas missense variants were concentrated in Pocket B. Among 1539 germline variants with known tumor laterality, 81. 2% were bilateral. Only 39. 4% of germline variants were reported in ClinVar. PANORAMA allowed interactive visualization and filtering of aggregated and user-uploaded RB1 variants. Conclusions: This work consolidates dispersed RB1 variant data into an unified resource and reveals domain-specific patterns of genetic disruption in RB. Integration with PANORAMA supports exploratory analysis, variant interpretation, and comparison of external cohorts to the global RB1 landscape.
Rooks et al. (Thu,) studied this question.