Shifting from DNA-Based to RNA-Informed Somatic Mutation Detection: A Systematic Benchmark Across Different Tumor Types

Accurate identification of somatic mutations is fundamental for cancer research. While their detection traditionally relies on DNA sequencing with matched tumor-normal samples, RNA sequencing (RNA-seq) offers distinct advantages by restricting analysis to expressed mutations with greater therapeutic potential. However, most RNA-seq datasets, particularly in clinical practice, lack matched normal samples, making reliable discrimination between somatic and germline variants a major challenge. As a result, RNA-based somatic mutation discovery in tumor-only settings remains inconsistent, with substantial variability across existing variant-calling pipelines and no established standard. This study systematically benchmarks somatic variant calling strategies for tumor-only RNA-seq data. Our analysis provides practical guidance for identifying expressed mutations, transforming how transcriptomic information can be leveraged for molecular characterization in precision oncology.