FVC as an adaptive and accurate method for filtering variants from popular NGS analysis pipelines

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Accurate quality control of genetic variants from whole-genome sequencing (WGS) is crucial for clinical diagnosis and human genetics research.
• Existing variant filtering methods have limitations, often removing more true variants than false ones and being specific to particular variant callers.

Purpose of the Study:

• To develop and validate an adaptive method for genetic variant quality control applicable across different analysis pipelines.
• To improve the accuracy of variant filtering by reducing false positives and minimizing the loss of true variants.

Main Methods:

• An adaptive quality control method, termed FVC (Filtering Variants Correctly), was developed.
• FVC was validated on genetic variants generated from four popular variant callers: GATK HaplotypeCaller, Mutect2, Varscan2, and DeepVariant.

Main Results:

• FVC demonstrated superior performance compared to existing state-of-the-art filtering methods.
• The method effectively removed a significantly higher number of false variants.
• FVC recalled approximately 51-99% of true variants that were filtered out by other methods.

Conclusions:

• FVC offers a robust and adaptive solution for genetic variant quality control, outperforming current methods.
• The trained FVC model can be easily integrated into user-specific variant calling pipelines for enhanced accuracy.