Computational phenotyping and genotype-phenotype association of STX1B-related epilepsy using the human phenotype ontology

<h3></h3> Characterisation of <i>STX1B</i>-related epilepsy is limited to small case series, single reports, and superficial cohort descriptions. We searched the literature for people with <i>STX1B</i> variants and seizures, and harmonised and re-analysed multisource genotypic and clinical data under a standardised phenotypic model using Human Phenotype Ontology and computational statistics to gain a comprehensive perspective including formal genotype- phenotype analysis. We found published information on 66 individuals. 68% individuals had generalised-onset seizures, 27% had focal-onset. 65% had motor seizures. 30% had treatment-resistant seizures. 64% had febrile seizures. Commonest seizures were bilateral tonic clonic (53% individuals), atonic (26%), absence (24%), myoclonic (24%). 47% had neurodevelopmental abnormalities: delay (26%), intellectual disability (11%), regression (9%), autism (6%). Two- sided Fisher’s exact tests showed significant genotype-phenotype associations for variant groups and specific protein variants. Nonsense variants (27%) associated with EEG spike- wave complexes, inversely with neurodevelopmental abnormalities. Missense variants (21%) associated with EEG focal discharges, ataxia. Deletions (9%) associated with neurodevelopmental phenotypes, language delay, facial dysmorphologies. Analysis of explicitly absent phenotypes and treatment outcomes continues. Our analysis elucidates the range of seizures, associated neurodevelopmental traits, and genotype-phenotype associations, comprising <i>STX1B</i>-related epilepsy. Accurate determination of fundamental phenotypic entities, medication responses and genotype- phenotype associations, will refine classification and guide precision care.