Aryana-bs: context-aware alignment of bisulfite-sequencing reads

Area of Science:

• Genomics
• Epigenetics
• Bioinformatics

Background:

• DNA methylation is crucial for biological processes and disease, with bisulfite sequencing (BS) as the gold standard for its measurement.
• Conventional short-read aligners struggle with BS-induced C-to-T conversions, leading to reduced accuracy.
• Existing alignment methods like three-letter and wildcard alignment have limitations, causing data loss or biased results.

Purpose of the Study:

• To develop a novel bisulfite sequencing aligner, ARYANA-BS, that directly integrates BS-specific base alterations.
• To enhance the accuracy of DNA methylation quantification by addressing challenges in read alignment.
• To provide a more robust tool for analyzing DNA methylation patterns across diverse genomic contexts.

Main Methods:

• ARYANA-BS constructs five indexes from the reference genome, leveraging known DNA methylation patterns.
• Each read is aligned to all indexes, with the alignment yielding the minimum penalty selected.
• An optional Expectation-Maximization (EM) step refines accuracy by integrating methylation probability information.

Main Results:

• ARYANA-BS demonstrates state-of-the-art accuracy in experimental evaluations on simulated and real data.
• The aligner maintains competitive speed and memory efficiency.
• It shows superior performance compared to BSMAP, bwa-meth, Bismark, BSBolt, and abismal.

Conclusions:

• ARYANA-BS significantly improves bisulfite sequencing data alignment accuracy by integrating methylation-specific alterations and genomic context.
• The aligner is robust against genomic biases and adept at aligning longer, higher-error reads.
• ARYANA-BS is suitable for cancer research and cell-free DNA studies, with the EM algorithm offering a framework for future enhancements.