NEAT-genreads

NEAT-genreads simulates sequencing reads to produce realistic test datasets for validation and benchmarking of genome analysis methods.

Key Features:

• Realistic Data Simulation: Generates sequencing reads that capture biological variants and sequencing artifacts to mimic real-world samples.
• Model Learning: Learns models from specific datasets to parameterize error and variant profiles and increase realism of simulated reads.
• Tunable Parameters: Provides adjustable parameters that can be set manually or parameterized using real datasets to customize simulations.
• Variant Comparison and Evaluation Scripts: Includes scripts for variant comparison and tool evaluation to support benchmarking of analysis methods.

Scientific Applications:

• Method Validation and Benchmarking: Enables validation and benchmarking of genomic analysis methods using simulated datasets with known ground truth.
• Reference Dataset Generation: Produces synthetic reference datasets that reflect specific mutational landscapes for testing analytical pipelines.
• Privacy-Preserving Human Genome Studies: Provides simulated human genome data when real human genome datasets are unavailable or restricted by privacy concerns.

Methodology:

Learns models from input datasets and uses those models to generate simulated sequencing reads that reflect the mutational landscape of sample genomes and include sequencing artifacts.