smCounter2: an accurate low-frequency variant caller for targeted sequencing data with unique molecular identifiers

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Low-frequency DNA mutations are challenging to detect due to technical artifacts in sample preparation and sequencing.
• Unique molecular identifiers (UMIs) correct most sequencing errors but cannot address pre-UMI tagging errors like DNA polymerase errors.
• These pre-UMI errors limit the accuracy of UMI-based variant calling.

Purpose of the Study:

• To develop an improved UMI-based variant caller, smCounter2, for targeted sequencing data.
• To enhance the detection limit and accuracy of low-frequency somatic variant identification.
• To provide a user-friendly and robust tool for analyzing sequencing data.

Main Methods:

• Development of smCounter2, an upgraded UMI-based variant caller.
• Implementation of a statistical test to assess variant allele frequency against background error rates.
• Inclusion of novel repetitive region filters specifically designed for UMI data to improve accuracy in non-coding regions.
• Benchmarking against state-of-the-art variant calling methods using multiple datasets.

Main Results:

• smCounter2 achieves a lower detection limit of 0.5%, down from 1% in previous versions.
• Demonstrates improved overall accuracy, particularly in non-coding regions.
• Exhibits superior performance in detecting somatic variants compared to existing methods.
• Offers consistent thresholding for both deep and shallow sequencing, with easier usability via Docker.

Conclusions:

• smCounter2 represents a significant advancement in UMI-based variant calling for targeted sequencing.
• The tool effectively addresses limitations of previous methods, enhancing the detection of low-frequency somatic mutations.
• smCounter2 provides a more accurate and accessible solution for genomic variant analysis.