CENTRE

CENTRE predicts cell-type-specific enhancer-target interactions to infer regulatory connections between enhancers and target promoters for elucidating gene regulation in specific cell types.

Key Features:

• Cell-type-specific prediction: Predicts enhancer-target (enhancer-promoter) interactions specific to the cell type of interest.
• Target promoter identification: Identifies target promoters regulated by active enhancers.
• Minimal input requirements: Uses gene expression data and ChIP-seq data for three histone modifications as the primary cell-type-specific inputs.
• Integration of external statistics: Incorporates cell-type-agnostic statistics derived from extensive datasets in public repositories alongside cell-type-specific data.
• Machine learning model: Employs a gradient boosting algorithm for prediction.
• Cross-dataset validation: Demonstrated performance matching or exceeding algorithms that require large-scale experimental data across multiple datasets and cell types.

Scientific Applications:

• Gene regulation studies: Infers enhancer-promoter relationships to aid investigation of transcriptional regulation mechanisms.
• Regulatory landscape mapping: Maps cell-type-specific regulatory interactions across different cell types.
• Disease and phenotype research: Links enhancers to target genes to support studies of regulatory variants implicated in phenotypes and diseases.
• Developmental biology: Resolves cell-type-specific regulatory interactions relevant to development.
• Oncology: Identifies regulatory interactions that may underlie cancer-related gene expression changes.
• Precision medicine: Supports interpretation of regulatory mechanisms in individualized genomic contexts.

Methodology:

Integrates gene expression data and ChIP-seq for three histone modifications with cell-type-agnostic statistics from public repositories and applies a gradient boosting algorithm to predict enhancer-target (enhancer-promoter) interactions.