Peak-agnostic high-resolution cis-regulatory circuitry mapping using single cell multiome data.

Area of Science:

• Genomics
• Molecular Biology
• Computational Biology

Background:

• Single-cell multiome sequencing (scRNAseq/ATACseq) offers high-resolution insights into cell-type specific transcriptional regulatory circuitry.
• Existing methods often fail to identify functional regulatory elements located outside of traditional chromatin 'peaks'.

Purpose of the Study:

• To introduce CREMA, a novel computational framework for reconstructing comprehensive cis-regulatory circuitry from single-cell multiome data.
• To overcome limitations of current approaches by identifying regulatory elements beyond called peaks.
• To demonstrate the utility of CREMA in uncovering cell-type specific regulatory mechanisms and disease-associated gene circuits.

Main Methods:

• CREMA models gene expression and chromatin activity within individual cells.
• The framework operates without requiring peak-calling or predefined cell type labels.
• It analyzes cis-regulatory elements both within and outside of chromatin peaks.

Main Results:

• CREMA successfully identifies functional regulatory elements missed by conventional methods, including those outside called peaks.
• These newly identified regulatory sites are crucial for distinguishing individual cell types.
• A Gata2-circuit regulating the Pcsk1 gene in mouse pituitary gonadotropes was identified and experimentally validated.

Conclusions:

• CREMA provides a more complete map of cis-regulatory circuitry than existing methods.
• Regulatory elements outside of chromatin peaks play significant cell-type specific roles.
• The framework and associated resources (human immune cell data, R package) facilitate further research into gene regulation.