Comprehensive benchmarking of somatic mutation detection by the SMaHT Network.

Using donor-specific and pangenome-inferred diploid assemblies improves somatic mutation calling, extending catalogs to challenging genomic regions in human tissues.

• Why it matters: Accurate detection of somatic mutations is crucial for understanding human biology and disease, but remains difficult due to technical limitations and complex genomic regions.
• What they did: The SMaHT Network conducted large-scale benchmarking with over 1,000× short-read and 100-400× long-read data across nine samples, evaluating sequencing, experimental, and computational methods, including six duplex technologies and single-cell sequencing.
• The result: Combining bulk, single-cell, and duplex analyses offers a comprehensive approach to characterize tissue mosaicism, providing a practical roadmap for accurate genome-wide somatic mutation detection in research and clinical settings.