ExoDiversity

ExoDiversity analyzes ChIP-exo data to deconvolve protein–DNA footprints and motifs, characterizing distinct transcription factor (TF) binding modes and their relationships to DNA shape and conservation.

Key Features:

• Model-Based Framework: Learns a joint distribution over footprints and motifs from ChIP-exo data without requiring prior motif or TF information and can adaptively determine the number of distinct binding modes.
• Resolution of Diverse Binding Modes: Decomposes mixed ChIP-exo footprints into distinct binding modes, distinguishing direct TF–DNA contacts and interactions mediated by intermediary proteins.
• Discovery of Co-factor Motifs and Variations: Identifies low-frequency co-factor TF motifs and detects small nucleotide variations within canonical motifs and surrounding regions associated with altered footprint patterns.
• Insights into DNA Shape and Conservation: Correlates detected binding modes with DNA shape features and sequence conservation signals to infer structural and functional aspects of TF–DNA interactions.

Scientific Applications:

• Cross-factor and cross-species analysis: Applicable to analysis of diverse transcription factors, organisms, and cell types using ChIP-exo data.
• Gene regulation and epigenetics: Enables investigation of complex regulatory mechanisms and heterogeneous binding behaviors relevant to gene regulation and epigenetic studies.
• Structural inference: Supports inference of alternative structural configurations of TF–DNA complexes by integrating footprint, motif, DNA shape, and conservation signals.

Methodology:

Processes ChIP-exo data to learn a joint distribution over footprints and motifs, separates distinct binding modes, and operates without prior motif or TF knowledge while adaptively determining the number of binding modes.