Integration of intra-sample contextual error modeling for improved detection of somatic mutations from deep...

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Sensitive mutation detection is crucial for early cancer detection, minimal residual disease (MRD) monitoring, and precision oncology.
• Artifacts in library preparation and sequencing complicate the specific detection of low-frequency variants.

Purpose of the Study:

• To introduce Espresso, a novel error suppression method for accurate single-nucleotide variant (SNV) detection.
• To evaluate Espresso's performance against existing advanced error suppression techniques.

Main Methods:

• Espresso utilizes local sequence features to enhance the accuracy of SNV detection.
• Performance comparison with other advanced error suppression techniques was conducted.

Main Results:

• Espresso demonstrated lower false-positive mutation calls and higher sensitivity compared to other methods.
• The method showed superior performance in detecting MRD in acute myeloid leukemia (AML) patients.
• Accurate mutation calling in limited genomic loci was shown to be effective for AML risk prediction.

Conclusions:

• Espresso offers a more sensitive and specific approach to mutation detection in next-generation sequencing data.
• The method has significant implications for MRD monitoring, precision oncology, and early cancer detection.
• Espresso's application extends to diverse research and clinical settings for accurate mutation identification.