Integrated analysis of multimodal long-read epigenetic assays

Area of Science:

• Epigenetics and Genomics
• Bioinformatics and Computational Biology

Background:

• Long-read sequencing assays are crucial for studying epigenetic regulation.
• Methods like DiMeLo-seq enable multimodal data generation, capturing multiple genomic signals on single molecules.
• Limited tools exist for visualizing and analyzing this complex multimodal data.

Purpose of the Study:

• Introduce dimelo-toolkit, a Python package for flexible visualization and analysis of multimodal long-read sequencing data.
• Demonstrate the utility of dimelo-toolkit with an extended DiMeLo-seq protocol for simultaneous chromatin state analysis.
• Facilitate the development of optimized protocols and unlock new biological insights.

Main Methods:

• Development of dimelo-toolkit, a Python package for data visualization and processing.
• Extension of the DiMeLo-seq protocol to capture three chromatin state aspects: protein binding, CpG methylation, and chromatin accessibility.
• Application of the multimodal method to map these features in GM12878 cells and investigate technical biases.

Main Results:

• dimelo-toolkit provides preset visualizations for multimodal base modification data.
• Simultaneous mapping of chromatin accessibility, CpG methylation, and LMNB1/CTCF binding patterns was achieved.
• Technical biases in multimodal data analysis were investigated using the toolkit.

Conclusions:

• dimelo-toolkit enhances the analysis of multimodal long-read sequencing data.
• The developed multimodal method and toolkit enable comprehensive chromatin state mapping.
• This software facilitates protocol optimization and discovery in epigenetics research.