Abstract:SCN2A -related disorders secondary to altered function in the voltage-gated sodium channel Na V 1.2 are rare with clinically heterogeneous expressions that include epilepsy, autism, and multiple severe to profound impairments and other conditions. To advance understanding of the clinical phenotypes and their relation to channel function, 81 patients (36, 44% female, median age 5.4 years) with 69 unique SCN2A variants were systematically phenotyped and their Na V 1.2 channel function systematically assessed. Participants were recruited through the FamileSCN2A Foundation. Primary phenotype (epilepsy of neonatal-onset, N=27; infant onset, N=18; and later onset N=24; and autism without seizures, (N=12) was strongly correlated with a non-seizure severity index (p=0.002), which was based on presence of severe impairments in gross motor, fine motor, communication abilities, gastrostomy tube dependence, and diagnosis of cortical visual impairment and scoliosis. Non-seizure severity was greatest in the neonatal-onset group and least in the autism group (p=0.002). Children with the lowest severity indices were still severely impaired, as reflected by an average Vineland adaptive behavior composite score of 49.5 (>3 SD below the test's norm-referenced mean). Epileptic spasms were significantly more common in infant onset (67%) than in neonatal (22%) or later-onset (29%) epilepsy (p=0.007). Primary phenotype also strongly correlated with variant function (p<0.0001); gain of function and mixed function variants predominated in neonatal-onset epilepsy, shifting to moderate loss of function in infant-onset epilepsy, and severe and complete loss of function in later-onset epilepsy and autism groups. Exploratory cluster analysis identified five groups representing (1) primarily later-onset epilepsy with moderate loss of function variants and low severity indices, (2) mostly infant-onset epilepsy with moderate loss of function variants but higher severity indices, (3) late-onset and autism only with the lowest severity indices (mostly 0) and severe/complete loss of function variants. Two exclusively neonatal clusters were distinguished from each other largely on non-seizure severity scores and secondarily on variant function. The relation between primary phenotype and variant function emphasizes the role of developmental factors in the differential clinical expression of SCN2A variants based on their effects on Na V 1.2 channel function. The non-seizure severity of SCN2A disorders depends on a combination of the age at seizure onset (primary phenotype) and variant function. As precision therapies for SCN2A -related disorders advance toward clinical trials, knowledge of the relationship between variant function and clinical disease expression will be valuable for identifying appropriate patients for these trials and in selecting efficient clinical outcomes.

Machine learning models reveal distinct disease subgroups and improve diagnostic and prognostic accuracy for individuals with pathogenic SCN8A gain-of-function variants

Patients carrying pathogenic SCN8A variants with loss‐ and gain‐of‐function effects can be classified into five subgroups exhibiting varying developmental and epileptic components of encephalopathy

Diagnostic Classification and Prognostic Prediction Using Common Genetic Variants in Autism Spectrum Disorder: Genotype-Based Deep Learning

Discordance between a deep learning model and clinical-grade variant pathogenicity classification in a rare disease cohort

Novel Association Strategy with Copy Number Variation for Identifying New Risk Loci of Human Diseases

Clinical Utility of Protein Language Models in Resolution of Variants of Uncertain Significance in KCNQ1, KCNH2, and SCN5A Compared With Patch-Clamp Functional Characterization

Development and Validation of a Prediction Model for Early Diagnosis of SCN1A-Related Epilepsies

Multi-view singular value decomposition for disease subtyping and genetic associations

Enhancing patient stratification and interpretability through class-contrastive and feature attribution techniques

Gene-SGAN: a method for discovering disease subtypes with imaging and genetic signatures via multi-view weakly-supervised deep clustering

Enhanced Classification of Brugada Syndrome-Associated and Long-QT Syndrome-Associated Genetic Variants in the SCN5A-Encoded Na(v)1.5 Cardiac Sodium Channel.

Gene-SGAN: discovering disease subtypes with imaging and genetic signatures via multi-view weakly-supervised deep clustering

Machine learning enabled subgroup analysis with real-world data to inform clinical trial eligibility criteria design

Machine learning modeling identifies hypertrophic cardiomyopathy subtypes with genetic signature

Classification of genetic variants using machine learning

Model-based clustering for identifying disease-associated SNPs in case-control genome-wide association studies

Expanded clinical phenotype spectrum correlates with variant function in SCN2A-related disorders

Resolution of SLC6A1 variable expressivity in a multi-generational family using deep clinical phenotyping and Drosophila models

Prognostic enrichment for early-stage Huntington's disease: An explainable machine learning approach for clinical trial

Machine Learning Study of SNPs in Noncoding Regions to Predict Non-small Cell Lung Cancer Susceptibility

Insights into clinical phenotypes and treatment responses in a Small cohort of Taiwanese patients with SCN1A variants: A Preliminary study