scANANSE gene regulatory network and motif analysis of single-cell clusters

Abstract

The recent development of single-cell techniques is essential to unravel complex biological systems. By measuring the transcriptome and the accessible genome on a single-cell level, cellular heterogeneity in a biological environment can be deciphered. Transcription factors act as key regulators activating and repressing downstream target genes, and together they constitute gene regulatory networks that govern cell morphology and identity. Dissecting these gene regulatory networks is crucial for understanding molecular mechanisms and disease, especially within highly complex biological systems. The gene regulatory network analysis software ANANSE and the motif enrichment software GimmeMotifs were both developed to analyse bulk datasets. We developed scANANSE, a software pipeline for gene regulatory network analysis and motif enrichment using single-cell RNA and ATAC datasets. The scANANSE pipeline can be run from either R or Python. First, it exports data from standard single-cell objects. Next, it automatically runs multiple comparisons of cell cluster data. Finally, it imports the results back to the single-cell object, where the result can be further visualised, integrated, and interpreted. Here, we demonstrate our scANANSE pipeline on a publicly available PBMC multi-omics dataset. It identifies well-known cell type-specific hematopoietic factors. Importantly, we also demonstrated that scANANSE combined with GimmeMotifs is able to predict transcription factors with both activating and repressing roles in gene regulation.

Cardiac differentiation roadmap for analysis of plasticity and balanced lineage commitment

“…To understand the transcriptional regulation driving CM specification in these stages of cardiac development, we predicted the CM-specific gene regulatory networks (GRNs) utilizing scANANSE, a recently developed tool for identification and prioritization of key transcription factors involved in cell fate determination (Smits et al, 2023; Xu et al, 2021). In short, scANANSE prioritizes transcription factors by integration of chromatin accessibility (sequence motif content) and transcriptome data, inferring a differential network between two biological states.…”

“…The resulting set was visualized in a bandplot using fluff (REF XX) and used as input for motif scanning analysis using the maelstrom function in Gimme Motifs (Bruse & van Heeringen, 2018). Motifs with in at least one of the clusters a z-score > 3, were selected and visualized with its potential binding transcription factors using an adjusted version of the Motif2TF function of scANANSE (Smits et al, 2023).…”

“…The Seurat implementation of scANANSE (AnanseSeurat, v1.1.0) was used to obtain the important transcription factor regulators from the Multiome dataset (Smits et al, 2023). The resulting differential network for the aCM cluster over vCM cluster, was filtered on the top 20 transcription factors with the highest influence score.…”