A review of somatic single nucleotide variant calling algorithms for next-generation sequencing data

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Somatic mutation detection is crucial for cancer treatment and research.
• Next-generation sequencing has spurred the development of numerous variant calling pipelines.
• Existing variant callers vary in models, filters, data needs, and applications.

Purpose of the Study:

• To provide a practical guide for selecting appropriate variant calling pipelines.
• To enumerate the unique features of state-of-the-art variant callers.
• To review benchmarking studies of variant callers.

Main Methods:

• Focus on somatic single nucleotide variant detection.
• Comparison of traditional whole genome/exome sequencing callers with low-frequency callers using unique molecular identifiers.
• Review of benchmarking materials, datasets, and performance metrics.

Main Results:

• Variant callers exhibit inconsistent performance across different benchmarking studies.
• A wide range of variant calling algorithms exist, from established to novel.
• Benchmarking methodologies and results vary significantly.

Conclusions:

• Selecting the right variant caller is critical for accurate somatic mutation detection.
• Understanding the nuances of different pipelines is essential for effective cancer research.
• Emerging trends and future directions in variant calling algorithms warrant further investigation.