Molecular mechanisms reconstruction from single-cell multi-omics data with HuMMuS.

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Cellular molecular identity arises from complex interactions across regulatory layers.
• Single-cell sequencing technologies enable the measurement of these regulatory layers.

Purpose of the Study:

• To introduce HuMMuS, a novel computational method for inferring regulatory mechanisms from single-cell multi-omics data.
• To enhance the analysis of regulatory networks by capturing macromolecular cooperation and integrating multiple molecular layers.

Main Methods:

• HuMMuS was developed to analyze single-cell multi-omics datasets.
• The method was benchmarked against state-of-the-art approaches using both paired and unpaired multi-omics data.
• HuMMuS was applied to single-nucleus methylome sequencing (snmC-seq), single-cell assay for transposase-accessible chromatin sequencing (scATAC-seq), and single-cell RNA sequencing (scRNA-seq) data from mouse brain cortex.

Main Results:

• HuMMuS demonstrated improved prediction accuracy for transcription factor (TF) targets, TF binding motifs, and regulatory regions compared to existing methods.
• Application to mouse brain multi-omics data resulted in accurate clustering of single-cell RNA sequencing profiles.
• The method successfully identified potential driver TFs regulating cellular states.

Conclusions:

• HuMMuS offers a powerful approach for dissecting regulatory mechanisms from single-cell multi-omics data.
• The method's ability to capture macromolecular cooperation and integrate diverse regulatory layers advances the field of systems biology.
• HuMMuS provides valuable insights into cell type classification and the identification of key regulatory factors in complex biological systems.