TechPrufer
Sign inGet started

Blogs›Cerebras›Clinical Variant Interpretation Benchmark

Clinical Variant Interpretation Benchmark

Clinical Variant Interpretation Benchmark

1
posts
2026

ExomeBench is a new benchmark for clinical variant interpretation in exome regions, designed to help researchers evaluate and improve models for health-relevant predictions. It includes five supervised classification tasks derived from ClinVar, focusing on pathogenicity, phenotype association, and gene attribution. The benchmark provides a standardized, reproducible dataset with baseline results and experimental artifacts to support transparent evaluation.

2026

Cerebras

2/23/2026

Introduced ExomeBench, a new benchmark for clinical variant interpretation. This involved curating over 158k single-nucleotide variants from ClinVar into five distinct classification tasks (pathogenicity, cancer-predisposing syndromes, cardiovascular phenotypes, BRCA classification, and top five genes). The post details the dataset structure, task definitions, data splitting strategies (including gene-disjoint splits for pathogenicity), and baseline performance metrics (MCC). It also provides resources for data loading, training, and evaluation.