Variant profiling of evolving prokaryotic populations.

Area of Science:

• Genomics
• Bioinformatics
• Microbiology

Background:

• Bacterial genomic heterogeneity is crucial in evolution and diagnostics.
• Short sequencing reads limit accurate detection of low-frequency variants.
• Next-generation sequencing (NGS) enables deep sequencing of bacterial populations.

Purpose of the Study:

• To develop a reliable workflow (VarCap) for predicting genomic variants, including single nucleotide polymorphisms (SNPs), insertions/deletions (InDels), and structural variations, even at low frequencies.
• To address challenges of false positive variant predictions caused by sequencing errors and mapping artifacts.
• To facilitate quantitative assessment of genomic heterogeneity in bacterial populations.

Main Methods:

• Evaluated sensitivity and accuracy of various software tools using synthetic read data.
• Developed VarCap workflow incorporating post-filtering of predicted variants based on frequency, coverage, and genomic context.
• Utilized an intersection of at least two tools for variant prediction to enhance precision.

Main Results:

• Achieved reliable prediction of variants with a minimum relative abundance of 2%.
• VarCap workflow demonstrated improved accuracy by filtering variants based on multiple properties.
• Sensitivity was maximized by combining different tools, while precision was enhanced by intersecting predictions from multiple tools.

Conclusions:

• VarCap provides a robust method for accurately capturing bacterial genomic heterogeneity from NGS data.
• The workflow is suitable for routine use in experimental evolution and clinical diagnostics.
• VarCap is available as open-source code and via a Galaxy webserver for broad accessibility.