phastSim: efficient simulation of sequence evolution for pandemic-scale datasets.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomic Epidemiology

Background:

• Sequence simulators are vital for bioinformatics tool development and inference methods.
• The rapid increase in genomic data, particularly SARS-CoV-2 genomes, challenges existing simulation software capabilities.
• Simulating large datasets efficiently is critical for advancing genomic research.

Approach:

• Developed a novel algorithm based on the Gillespie approach for efficient sequence evolution simulation.
• Implemented a multi-layered search tree structure to optimize computational performance.
• Leveraged the biological principle that only a small fraction of a genome mutates per branch.

Key Points:

• The software efficiently simulates sequence evolution along extremely large phylogenetic trees (>100,000 tips) with short branches.
• Achieves high computational efficiency by exploiting limited mutation rates per branch.
• Offers flexibility with various evolutionary models, including indel and novel hypermutability models for SARS-CoV-2.

Conclusions:

• The new open-source software (phastSim) addresses the limitations of current sequence simulators for massive datasets.
• Facilitates easier integration with other Python packages for broader bioinformatics applications.
• Provides a more realistic simulation of viral genome evolution, particularly for pathogens like SARS-CoV-2.