BINDER

BINDER infers gene regulatory networks in Mycobacterium abscessus by integrating gene coexpression and comparative genomics with RNA-seq primary data and auxiliary ChIP-seq data to probabilistically identify regulator-target interactions.

Key Features:

• Hybrid Approach: Combines gene coexpression data with comparative genomics and genomic conservation properties to inform regulatory inference.
• Data Integration: Integrates primary RNA-seq data and sequence information from Mycobacterium abscessus with auxiliary ChIP-seq data from Mycobacterium tuberculosis.
• Hierarchical Bayesian Framework: Implements a hierarchical Bayesian model that informs bivariate likelihood functions and prior distributions to combine primary and auxiliary data.
• Probabilistic Inference: Performs probabilistic scoring of regulator-target pairs and identified 54 significant pairs among 167,280 potential regulator-target pairs across five transcription factors.

Scientific Applications:

• Regulon Grouping Insight: Provides insights into regulon groupings and transcriptional control mechanisms in Mycobacterium abscessus.
• Broader Applicability: Applies to other organisms where integration of primary data and proxy auxiliary data (for example related-species ChIP-seq) is required for computational inference of gene regulatory networks.

Methodology:

Uses RNA-seq data from Mycobacterium abscessus and sequence information as primary data; incorporates ChIP-seq data from Mycobacterium tuberculosis as auxiliary data; applies a hierarchical Bayesian model that informs bivariate likelihood and prior distributions; performs probabilistic scoring of regulator-target pairs to infer regulatory interactions.