TimelessFlex

TimelessFlex analyzes chromatin state trajectories in promoter-enhancer pairs by integrating time series Hi-C, ChIP-seq, and ATAC-seq data to characterize temporal regulatory mechanisms during tissue differentiation.

Key Features:

• Integration with Hi-C Data: Utilizes time series Hi-C to connect promoters and enhancers and track changes in chromatin contact frequencies across multiple time points.
• Co-clustering of Histone Modifications: Applies a Bayesian network approach to co-cluster multiple histone modification ChIP-seq datasets at promoter and enhancer regions to define chromatin states.
• Expectation-Maximization Algorithm: Employs an expectation-maximization algorithm to assign promoters and enhancers based on Hi-C interactions and to jointly cluster feature regions.
• Dynamic Chromatin Accessibility Analysis: Uses time series ATAC-seq to define candidate promoters and enhancers and to measure temporal changes in open chromatin states.
• Correlation with Gene Expression: Relates promoter cluster patterns to gene expression signals and evaluates trends in Hi-C signal changes relative to activation.

Scientific Applications:

• Tissue Differentiation Studies: Characterizes promoter-enhancer chromatin state trajectories in linear and branched tissue differentiation processes.
• Epigenetics and Developmental Biology: Identifies temporal chromatin-state changes and regulatory mechanisms relevant to developmental gene regulation.
• Disease Progression and Therapeutic Research: Supports analysis of chromatin modification dynamics relevant to disease progression and interventions targeting chromatin states.

Methodology:

Uses time series ChIP-seq for histone modifications (≥ three time points); defines promoter and enhancer candidates using time series ATAC-seq; integrates time series Hi-C to assign promoter-enhancer links; and applies a Bayesian network-based co-clustering approach together with an expectation-maximization algorithm to jointly cluster feature regions into paired chromatin state trajectories.