wGRN

wGRN reconstructs genome-wide gene regulatory networks in wheat by integrating large-scale functional datasets to elucidate regulation of phenotypic traits and prioritize candidate genes.

Key Features:

• Integration of Diverse Functional Datasets: Integrates gene expression profiles, sequence motifs, transcription factor (TF) binding data, chromatin accessibility information, and evolutionarily conserved regulatory elements to inform regulatory relationships.
• Extensive Interaction Mapping: Represents 7.2 million genome-wide interactions involving 5,947 TFs and 127,439 target genes.
• Interaction Verification: Verifies interactions using known regulatory relationships, condition-specific expression patterns, functional annotations, and experimental validations.
• Pathway Assignment and Gene Prioritization: Assigns over 12,000 genes to biological pathways and prioritizes candidate genes identified in genome-wide association studies (GWAS).
• Temporal Transcriptome Network Construction: Constructs high-resolution networks from spike temporal transcriptome datasets to capture regulatory dynamics over time.
• Machine Learning for Regulator Identification and Trait Prediction: Applies machine learning to identify novel regulators and improve prediction accuracy of spike phenotypic traits, aiding interpretation of phenotypic differences among modern wheat accessions.

Scientific Applications:

• Trait-Associated Gene Discovery: Links functional datasets and regulatory interactions to identify and prioritize genes associated with crop traits.
• Crop Improvement Strategies: Provides pathway- and regulator-level targets for breeding efforts aimed at yield, resilience, and quality improvements.
• Understanding Phenotypic Variability: Enables analysis of regulatory mechanisms underlying phenotypic differences among wheat varieties, including modern accessions.

Methodology:

Constructs regulatory networks by integrating gene expression, sequence motifs, TF binding data, chromatin accessibility, and conserved regulatory elements; verifies interactions via known regulatory relationships, condition-specific expression patterns, functional annotations and experimental validations; comprises 7.2 million interactions (5,947 TFs, 127,439 targets); assigns >12,000 genes to pathways; analyzes spike temporal transcriptome datasets to build high-resolution networks; and applies machine learning to identify novel regulators and predict spike phenotypic traits.