Multiscale footprints reveal the organization of cis-regulatory elements

Area of Science:

• Genomics
• Computational Biology
• Epigenetics

Background:

• Cis-regulatory elements (CREs) are crucial for gene expression but their dynamic protein organization is poorly understood.
• Existing methods limit genome-wide analysis of effector protein organization at CREs, hindering functional studies in cell fate and disease.

Purpose of the Study:

• To develop a computational method for measuring effector protein organization at CREs across the genome.
• To infer transcription factor and nucleosome binding and interpret regulatory logic at CREs.
• To investigate CRE dynamics during hematopoiesis and aging.

Main Methods:

• Developed PRINT, a computational method to identify DNA-protein interaction footprints from chromatin accessibility data.
• Created the seq2PRINT framework using deep learning for precise inference of protein binding.
• Applied seq2PRINT to single-cell chromatin accessibility data from human bone marrow and murine hematopoietic stem cells.

Main Results:

• Observed sequential establishment and widening of CREs centered on pioneer factors during human hematopoiesis.
• Discovered age-associated alterations in CRE structure in murine hematopoietic stem cells, including reduced nucleosome footprints.
• Identified a gain of de novo Ets composite motifs in aged murine hematopoietic stem cells.

Conclusions:

• Established a method to gain insights into DNA-binding protein dynamics from chromatin accessibility data.
• Revealed the architecture of regulatory elements across differentiation and aging processes.
• Provided a framework for understanding regulatory element function in cell fate and disease.