A dataset generator for whole genome shotgun sequencing.

Area of Science:

• Computational Biology
• Bioinformatics
• Genomics

Background:

• Simulated datasets are crucial for software development, enabling isolated study of phenomena and validation against known solutions.
• Developing robust software for assembling large-scale genomic data, such as whole human genome shotgun data, presents significant challenges.

Purpose of the Study:

• To introduce celsim, a novel simulator for generating realistic, complex DNA sequence data.
• To facilitate the development and testing of software for whole human genome shotgun data assembly.

Main Methods:

• Developed celsim, a simulator utilizing a stochastic grammar.
• Enabled description and stochastic generation of target DNA sequences with varied repeat structures.
• Incorporated functionality to generate polymorphic variants and shotgun sequencing data sets.

Main Results:

• Demonstrated the capability of celsim to generate complex DNA sequences and associated shotgun data.
• Highlighted the tool's simplicity in design despite its powerful simulation capabilities.
• Validated the utility of simulated data for studying specific genomic phenomena in isolation.

Conclusions:

• celsim is an invaluable and powerful tool for generating simulated whole human genome shotgun data.
• The simulator's design, based on stochastic grammar, allows for flexible and detailed data generation.
• This tool significantly aids in the development and benchmarking of genomic data assembly software.