Scholarlink

<p>Genotype-first screening compares patient genomes to a standard reference. The GRCh38 linear assembly contains reference minor alleles (RMAs), posing a critical structural vulnerability for automated clinical variant classification. To determine whether RMAs systematically generate false-positive annotations, we conducted an observational case series of 20 healthy adults undergoing preventive genome sequencing. Automated bioinformatic processing was performed using the GRCh38 linear reference to identify “high-impact” variant annotations generated at loci where GRCh38 differs from the population consensus major allele. Among all 20 participants (100%), linear alignment to GRCh38 systematically misclassified functional major alleles as false-positive “high-impact” annotations at three distinct loci (<i>SLC37A4</i>, <i>CIMIP2A</i>, and <i>GPR33</i>). These artifacts occurred solely because analysis software mathematically defined the healthy wild-type state as a deviation from the rare RMA. Cross-referencing with gnomAD confirmed these as population-dominant benign alleles. Furthermore, this inherent reference bias introduces a theoretical risk of false-negative classifications if RMAs mathematically mask true pathogenic variants. Automated pipelines using linear references systematically misclassify healthy alleles as high-impact functional annotations. While downstream population-frequency filters manage false-positive artifacts, this retrospective patching creates an unsustainable bottleneck for population-scale screening. Transitioning to graph-based pangenome references represents a highly promising approach to directly resolve these diagnostic vulnerabilities at the alignment level, though computational and standardization challenges must first be addressed to ensure the accuracy of large-scale aging research.</p> Graphical Abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Beyond the linear genome: how reference bias threatens preventive medicine and geroscience

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月25日

查看全文
  • Gyula Richárd Nagy,
  • Gyöngyi Munkácsy,
  • Ankita Murmu,
  • Giuliana Longo,
  • Luis Izquierdo López,
  • Vincenzo Cirigliano,
  • Balázs Győrffy

摘要

Genotype-first screening compares patient genomes to a standard reference. The GRCh38 linear assembly contains reference minor alleles (RMAs), posing a critical structural vulnerability for automated clinical variant classification. To determine whether RMAs systematically generate false-positive annotations, we conducted an observational case series of 20 healthy adults undergoing preventive genome sequencing. Automated bioinformatic processing was performed using the GRCh38 linear reference to identify “high-impact” variant annotations generated at loci where GRCh38 differs from the population consensus major allele. Among all 20 participants (100%), linear alignment to GRCh38 systematically misclassified functional major alleles as false-positive “high-impact” annotations at three distinct loci (SLC37A4, CIMIP2A, and GPR33). These artifacts occurred solely because analysis software mathematically defined the healthy wild-type state as a deviation from the rare RMA. Cross-referencing with gnomAD confirmed these as population-dominant benign alleles. Furthermore, this inherent reference bias introduces a theoretical risk of false-negative classifications if RMAs mathematically mask true pathogenic variants. Automated pipelines using linear references systematically misclassify healthy alleles as high-impact functional annotations. While downstream population-frequency filters manage false-positive artifacts, this retrospective patching creates an unsustainable bottleneck for population-scale screening. Transitioning to graph-based pangenome references represents a highly promising approach to directly resolve these diagnostic vulnerabilities at the alignment level, though computational and standardization challenges must first be addressed to ensure the accuracy of large-scale aging research.

Graphical Abstract