FIREVAT: finding reliable variants without artifacts in human cancer samples using etiologically relevant mutational...

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Accurate somatic variant identification is crucial for cancer genomics and precision oncology.
• Sequencing artifacts often compromise the confidence of variant calls.
• Existing computational methods for variant filtering are resource-intensive.

Purpose of the Study:

• To introduce FIREVAT, a novel software for efficient and accurate somatic variant refinement.
• To leverage mutational signatures of sequencing artifacts for improved variant filtering.

Main Methods:

• FIREVAT utilizes known artifact mutational signatures from large human tumor sample catalogs.
• The software refines variants by distinguishing artifactual mutations from real somatic mutations.
• Performance was validated using orthogonal sequencing datasets and compared against existing approaches.

Main Results:

• FIREVAT accurately removes artifacts, significantly enhancing the precision and specificity of somatic variant calls.
• The software achieved superior performance compared to current filtering methods.
• Application on TCGA data revealed more biologically relevant signatures and enriched error-associated contexts.

Conclusions:

• FIREVAT offers a novel strategy for refining somatic point mutations in cancer genomes using mutational signatures.
• The tool is expected to advance precision oncology by enabling more accurate variant identification and analysis of prognostic/therapeutic mutational signatures.
• FIREVAT is publicly available for use.