ComB: SNP calling and mapping analysis for color and nucleotide space platforms

Area of Science:

• Genomics and Bioinformatics
• Computational Biology
• Molecular Genetics

Background:

• Next-generation sequencing (NGS) platforms like SOLiD have accelerated SNP determination.
• Current SNP calling methods often translate color-space reads to nucleotide space, losing information and causing errors, particularly with dense or adjacent polymorphisms.
• Ambiguous color read alignments further complicate accurate SNP identification.

Purpose of the Study:

• To develop a novel SNP calling tool, ComB, that operates directly in color space.
• To address the limitations of existing methods in accurately calling SNPs from SOLiD sequencing data, especially in challenging polymorphic regions.
• To enable efficient and accurate whole-genome SNP calling using large-scale sequencing data.

Main Methods:

• Developed ComB, a SNP calling tool utilizing a Bayesian model operating directly on color-space reads.
• ComB incorporates unique and ambiguous reads through an iterative process to determine SNP identity.
• The tool recalibrates quality scores and can integrate both sequence and color-space data.

Main Results:

• ComB accurately identifies short consecutive nucleotide polymorphisms and densely clustered SNPs, outperforming existing tools.
• The tool demonstrates superior performance on both real and simulated data compared to nucleotide-space translation methods.
• ComB achieves higher true SNP discovery rates and lower false positive rates.

Conclusions:

• ComB offers a significant advancement in SNP calling accuracy and efficiency by leveraging color-space data directly.
• The Bayesian iterative strategy and quality score recalibration are key to ComB's improved performance.
• ComB enables accurate and parallel whole-human-genome SNP calling, even with billions of short reads.