Detecting expansion of short tandem repeat in whole exome sequencing data of movement disorder cohort with bioinformatics tool expansion hunter

Abstract

Background: Short tandem repeat (STR) expansions are associated with several neurological and neuromuscula disorders such as spinocerebellar ataxias. Expansion mutation of different STR loci causes overlapping clinical phenotypes, therefore a high-throughput screening tool is important. Whole exome sequencing (WES) has this advantage of profiling multiplex loci within one run when compared to conventional diagnostic assay, but it is challenging to accurately detect expansion changes with short read sequencing. This pilot study aims to assess the performance of ExpansionHunter, a recently developed bioinformatics tool that evaluates the sizes of repeats on 25 STR loci using short read sequencing data. Allele frequencies of these clinically important genes in Hong Kong population can be reference for further studies. Methods: ExpansionHunter (version 3.2.2) was firstly used on 759 WES data of a Hong Kong Chinese cohort (382 males and 377 females) that do not have indication of movement disorder. Allele frequencies of 25 STR loci was analysed with frequency distribution plots. ExpansionHunter was then used on the WES data of our 68 movement disorder proband cohort (43 males and 25 females). Outliers were identified by comparing the sizes of repeats called in their WES data with population allele frequencies, and severities were determined by referring to literatures. Results: ExpansionHunter identified exonic STR in WES data better than identifying STR located in untranslated region, promoter and intron. In particular, exonic repeats in AR, ATN1, ATXN1, CACNA1A, JPH3, and TBP were detected in 87.05%~100.00% of the cohort. The allele frequency distribution plots generally were unimodal, and outliers with expanded repeats were rare. ExpansionHunter successfully detected the positive control with known DMPK full mutation expansion in the movement disorder proband cohort, confirming its performance on WES data. Among the rest of the probands, one outlier was identified to have full mutation expansion in ATXN8OS and another one was identified to have full mutation expansion in TBP. Both outliers had negative diagnosis for SNV previously. Further investigation with alternative diagnostic methods will be done to validate the finding of ExpansionHunter. Conclusion: This study demonstrated the application of ExpansionHunter on WES data to identify potential STR expansions. Through the comparison of repeat numbers called in patient samples and the population STR variant frequencies, outliers that were in the range of full mutation penetrance were identified. By integrating this kind of STR genotyping tools into our bioinformatics pipeline, the spectrum of pathogenic variants detectable by WES can be expanded and diagnostic yield can be increased.