Computational short tandem repeat genotyping reveals clinically relevant expansions in a large Turkish neurodegeneration disease cohort

Abstract

Short tandem repeat (STR) expansions are a major cause of neurodegenerative disorders; however, their genetic and clinical heterogeneity complicates diagnosis. STR detection remains limited in routine short-read next-generation sequencing (NGS) workflows. We evaluated the diagnostic yield and clinical utility of computational STR genotyping in a large Turkish neurodegenerative disease cohort. ExpansionHunter was applied to NGS data from 3150 patients and 146 controls, targeting 15 disease-associated STR loci. To improve genotyping of poorly captured exonic regions in exome data, the default locus coverage threshold was reduced from 10× to 3×. Candidate expansions were visually inspected using REViewer and validated by conventional molecular methods. Computational analysis detected 28 pathogenic and 160 intermediate expansions. Of these, 23 were confirmed as pathogenic, and eight initially classified as intermediate were reclassified as pathogenic after conventional validation, resulting in 31 pathogenic cases across 28 families: HTT (n = 8), ATXN2 (n = 5), ATXN1 (n = 4), DMPK (n = 3), PABPN1 (n = 3), TBP (n = 2), and single cases in AR, ATN1, and CACNA1A. Lowering the coverage threshold markedly increased genotyping rates at low-coverage loci in exome data, particularly in ATXN2. Genetic findings were largely consistent with clinical pre-diagnosis and the additional diagnostic yield was 0.95%. These findings support integrating STR analysis into routine neurogenetic diagnostics.