AIRVF: a filtering toolbox for precise variant calling in Ion Torrent sequencing.

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Ion Torrent sequencing is widely used in healthcare research and industry.
• Platform-specific artifacts can lead to inaccurate variant identification, particularly for low-frequency variants (<15%).

Purpose of the Study:

• To develop a robust filtering toolbox, Artifact Identification and Reduction of Variant False positives (AIRVF), to minimize artifact-driven false variant calls.
• To enhance the accuracy of variant detection in Ion Torrent sequencing data.

Main Methods:

• Developed a multi-step filtering toolbox (AIRVF) operating on flowgram, raw reads, mapped reads, and called variants.
• Tested AIRVF on sequencing data from standard reference materials.

Main Results:

• AIRVF achieved up to a ~98% reduction in false variants when integrated with public pipelines.
• A ~48% reduction in false variants was observed when combined with a commercial in-house solution.
• Minimal loss of sensitivity was maintained during the filtering process.

Conclusions:

• AIRVF effectively reduces false variant calls in Ion Torrent sequencing data.
• The toolbox enhances variant accuracy, crucial for reliable healthcare research and diagnostics.
• AIRVF offers a valuable solution for improving the quality of sequencing data analysis.