Crop-GPA

Crop-GPA aggregates and analyzes gene-phenotype association (GPA) data in crop plants to support discovery and molecular breeding.

Key Features:

• Comprehensive Database: A curated repository of genes, phenotypes, and gene-phenotype associations in crop plants.
• Dynamic Visualizations: Graphical visualizations for exploring and interpreting complex gene-phenotype relationships.
• GPA-BERT: Automatically extracts gene-phenotype associations from bio-crop literature.
• GPA-GCN: Predicts unknown gene-phenotype relationships using graph convolutional networks (GCNs).

Scientific Applications:

• Literature-based GPA extraction: Systematic extraction of gene-phenotype associations from bio-crop literature to build association datasets.
• Prediction of candidate GPAs: Identification of putative gene-phenotype relationships based on existing data using GCNs.
• Molecular breeding and marker discovery: Support for identifying genetic markers and associations linked to traits such as yield, resilience, and nutritional value.

Methodology:

GPA-BERT automatically extracts gene-phenotype associations from bio-crop literature. GPA-GCN predicts unknown gene-phenotype relationships by applying graph convolutional networks (GCNs).