STRtr-Kit: A classification system and evidence criteria for tandem repeat-disease relationships
ACMG 2026
Laurel Hiatt, Pamela Ajuyah Robertson, Harriet Dashnow, Macayla Ann Weiner
Introduction
Tandem repeats (TRs)—including short tandem repeats (1-6 bp motifs) and variable number tandem repeats (7+ bp motifs)—have been linked to dozens of Mendelian diseases, such as Huntington’s, and several forms of ataxia and epilepsy. Current frameworks for evaluating gene-disease relationships do not adequately address TR-specific complexities. For example, TR diseases often show genetic anticipation, where the age of onset and severity are associated with increasing allele size. As a result, proposed TR-disease relationships are often incorrectly classified, under-evaluated, or excluded entirely, limiting discovery and hindering clinical application.
Methods
To address this gap, we developed STRtr-Kit, a standardized scoring framework designed to accurately evaluate TR-disease relationships. STRtr-Kit builds on current best-practice scoring procedures developed by ClinGen, by curating based on locus-level (rather than gene-level) classification, introducing TR-specific evidence categories, and reweighting scoring. We applied the STRtr-Kit framework to curate 75 loci from STRchive, a database of disease-associated TRs, scoring each based on publicly available genetic and experimental evidence.
Results
We compared the STRtr-Kit curations to all relevant gene-level curations submitted by 12 scoring groups of the Gene Curation Coalition, including ClinGen, Orphanet, Ambry Genetics, and Genomes2People. Of the 75 loci scored using STRtr-Kit, only 38 had been previously evaluated. The STRtr-Kit framework better captured the distinct features of TR pathogenicity, such as repeat motif and length consequences. Many well-established TR-disease associations occur at genes where no disease association had previously been reported. STRtr-Kit was able to curate these associations by allowing TR-specific evidence to drive scoring. This allowed us to score previously uncurated associations, bringing the new number of curated TR-disease associations to 72, and to identify three contradictory (disputed and refuted) associations.
Conclusion
By explicitly incorporating TR-specific evidence into the curation process, STRtr-Kit provides a clear and reproducible methodology for assessing TR locus-disease relationships. This framework aims to improve classification consistency across scorers and establish a foundation for fully integrating tandem repeats into the field of genetic medicine. This new procedure allowed us to curate previously unassessed TR-disease associations, supporting an increase in the number and accuracy of genetic diagnoses.
Related publications