Abstract
Short tandem repeats (STRs) are highly abundant in the human genome and their age-related somatic instability is emerging as a pivotal pathomechanism in repeat expansion disorders. Nevertheless, currently no analysis tool exists for genome-wide profiling of STR somatic instability. Here, we present searchSTR, a novel computational framework that enables accurate STR genotyping and somatic instability quantification from both whole-genome and targeted sequencing data. Applying this approach, we analyzed STR variants from whole-genome sequencing data of the 1000 Genomes Project (n=3,202) and targeted deep-sequencing data of the population-based Rhineland Study (n=2,974). We provide a comprehensive multiancestry genome-wide reference panel of STR variants and their somatic instability, covering more than 1.4 million STRs across 26 populations. Remarkably, STR somatic instability at many loci was robustly associated with age, sex, brain morphology and markers of neurodegeneration. These findings reveal a crucial role for STR somatic instability as an age-dependent modifier of brain structure.
Introduction
Short Tandem Repeats (STRs) are sequences of 1–6 base pairs that repeat consecutively throughout the human genome. While these repeats are inherited in a specific state (germline), they are inherently unstable. Throughout an individual's lifespan, these sequences can expand or contract during cell division, a biological phenomenon known as somatic instability.
While massive expansions in specific genes are well-documented causes of "repeat expansion disorders" such as Huntington’s disease, the broader impact of genome-wide somatic instability has remained largely unexplored. This gap is primarily due to the lack of computational tools capable of distinguishing true biological mosaicism from sequencing noise at a population scale. This study introduces searchSTR, a framework designed to move beyond single-gene analysis to provide a comprehensive profile of somatic instability across the entire genome.
Methodology: The searchSTR Framework
The researchers developed searchSTR, implemented within the STRbean package, to address the technical challenges of quantifying mosaicism. The framework introduces several key innovations:
- Non-Symmetric Modeling: Traditional tools often assume that repeat variations follow a symmetric distribution. However, biological expansions and contractions do not occur at equal rates. searchSTR utilizes a mixture of two skewed distributions within a Maximum Likelihood Estimation (MLE) framework to more accurately reflect biological reality.
- Computational Efficiency: By utilizing the SeqAn library and FM-indexing, searchSTR achieves high-speed processing, making it feasible to analyze massive datasets like the 1000 Genomes Project.
- Somatic Expansion Instability (SEI) Score: To standardize measurements, the team defined the SEI score, a metric that quantifies the degree of mosaicism at each STR locus, allowing for robust comparisons across individuals and populations.
Data and Population Profiling
The study leveraged two major cohorts to validate the tool and explore biological associations:
- 1000 Genomes Project (n=3,202): This dataset allowed for the creation of a multiancestry reference panel covering 1.4 million STRs across 26 global populations.
- The Rhineland Study (n=2,974): A population-based study providing targeted deep-sequencing data paired with extensive phenotyping, including brain MRI scans and blood-based biomarkers.
- Dutch Benchmark Cohorts: Used for validation against gold-standard PCR-based measurements.
Key Findings
Ancestry and Sex-Linked Instability
The analysis revealed that genomic instability is not uniform across populations. Individuals of African ancestry (AFR) exhibited the highest overall levels of genomic instability. Furthermore, while males showed higher overall instability, the researchers discovered this was almost exclusively driven by the Y chromosome. When focusing on autosomes, sex-based differences in instability were negligible.
A Genetic Clock for Brain Structure
One of the study's most significant findings is the robust association between somatic instability and age. Higher average SEI scores (avgSEI) were significantly linked to reduced volumes in the hippocampus, gray matter, and cerebellar cortex—regions typically associated with aging and neurodegenerative decline.
Specifically, the study identified 43 STR variants where instability was statistically linked to brain morphology. Notable examples include instability in the CACNA1A gene (associated with ataxia) and HOXD13.
Markers of Neurodegeneration
Beyond structural imaging, the study linked somatic instability to molecular markers of damage. Instability at the GULOP locus was positively associated with plasma levels of Neurofilament Light Chain (NfL), a critical biomarker for neuronal injury and neurodegeneration.
Performance and Benchmarking
In comparative tests, searchSTR significantly outperformed existing tools like ExpansionHunter. It demonstrated higher accuracy (measured by intraclass correlation with PCR ground truth) while being up to two orders of magnitude faster and requiring substantially less memory. This efficiency is vital for the transition toward large-scale clinical and population genomics.
Limitations and Future Directions
While searchSTR represents a significant advancement, the study notes that the use of "spanning reads" may limit the detection of extremely long expansions that exceed standard sequencing lengths. Additionally, because the DNA was sourced from peripheral blood, it may not fully capture the specific mosaicism occurring within brain tissue.
Future research will focus on integrating these STR maps with single-cell sequencing and epigenetic data to determine how these "genetic stutters" disrupt gene regulation and contribute to the progression of age-related neurological diseases.
Summary updated