Get started
Work

Implementation of a Multi-Tiered Missense Predictor Pipeline for Variant Prioritization in Sleep-Related Hypermotor Epilepsy

Year: 2026
Type: article
Abstract: Introduction: Epilepsy is a common neurological disorder characterized by recurrent seizures, in which genetic factors play a key role. Among epilepsy-associated genes, CHRNA2, CHRNA4, and CHRNB2 encode subunits of neuronal nicotinic acetylcholine receptors (nAChRs), which are essential for regulating neuronal excitability and synaptic transmission. However, many missense variants in these genes are classified as variants of uncertain significance (VUS), limiting their clinical interpretation. Methods: We evaluated missense VUS in CHRNA2, CHRNA4, and CHRNB2 using a gene-set-calibrated, mechanism-aware in silico framework. LoGoFunc assigned each candidate to a gain- or loss-of-function branch. We ranked both branches using a REVEL–AlphaMissense–CADD consensus layer, MutScore, and PrimateAI; we also used ConSurf, DynaMut2, and population rarity for the loss-of-function branch. We assessed performance against a curated benchmark set of 41 pathogenic or likely pathogenic and 136 benign or likely benign missense variants. Results: The framework identified 381 candidate variants, including 259 in the gain-of-function branch, 115 in the loss-of-function branch, with seven unassigned. Of these, 94 gain-of-function and 38 loss-of-function candidates met the highest-priority criteria. During candidate retrieval, we recovered 35 of 41 pathogenic reference variants. The fully supported tier achieved 97.8% specificity, 88.5% positive predictive value, and a Matthews correlation coefficient of 0.642. Conclusion: This mechanism-aware framework reduces a broad range of uncertain variants to a prioritized list of testable candidates, each with a distinct functional hypothesis for laboratory validation.
DOI: 10.1159/msy/adrag029
Language: English
Subfield: Genetics
Open Access status: green
OpenAlex ID: W7204208967
FWCI:
Cites:
Cited by:
Related to: 9