DeepBSA

DeepBSA performs bulked segregant analysis (BSA) on high-throughput sequencing data to map mutations and quantitative trait loci (QTLs) in animals and plants.

Key Features:

• Deep learning-driven methodology: Implements a deep learning-based algorithm tailored for BSA that reduces absolute bias and enhances the signal-to-noise ratio.
• Algorithm integration: Implements a k-value algorithm and integrates five widely used BSA algorithms alongside the two newly developed algorithms (deep learning and k-value).
• High-throughput sequencing support: Analyzes high-throughput sequencing data to improve accuracy in QTL mapping and functional gene cloning.
• Multiple bulk pools compatibility: Processes a variable number of bulked pools, supporting at least two bulked pools per analysis.
• Performance validation: Demonstrated superior accuracy and reduced error rates in comparative analyses on simulated and real datasets from animal and plant studies.

Scientific Applications:

• Functional genomics and breeding: Enables precise identification of QTLs and candidate genes associated with complex traits to support functional genomics studies and breeding programs.
• Empirical mapping example: Applied to an F2 segregating maize population of 7,160 individuals, identifying five candidate QTLs, including three known plant-height genes.

Methodology:

Implements a deep learning-based algorithm and a k-value algorithm, integrates five established BSA algorithms, and analyzes high-throughput sequencing data; validated by comparative analyses on simulated and real animal and plant datasets.