GRaNIE and GRaNPA: inference and evaluation of enhancer‐mediated gene regulatory networks

Abstract

Enhancers play a vital role in gene regulation and are critical in mediating the impact of noncoding genetic variants associated with complex traits. Enhancer activity is a cell‐type‐specific process regulated by transcription factors (TFs), epigenetic mechanisms and genetic variants. Despite the strong mechanistic link between TFs and enhancers, we currently lack a framework for jointly analysing them in cell‐type‐specific gene regulatory networks (GRN). Equally important, we lack an unbiased way of assessing the biological significance of inferred GRNs since no complete ground truth exists. To address these gaps, we present GRaNIE (Gene Regulatory Network Inference including Enhancers) and GRaNPA (Gene Regulatory Network Performance Analysis). GRaNIE (https://git.embl.de/grp-zaugg/GRaNIE) builds enhancer‐mediated GRNs based on covariation of chromatin accessibility and RNA‐seq across samples (e.g. individuals), while GRaNPA (https://git.embl.de/grp-zaugg/GRaNPA) assesses the performance of GRNs for predicting cell‐type‐specific differential expression. We demonstrate their power by investigating gene regulatory mechanisms underlying the response of macrophages to infection, cancer and common genetic traits including autoimmune diseases. Finally, our methods identify the TF PURA as a putative regulator of pro‐inflammatory macrophage polarisation.

Glycolysis–Wnt signaling axis tunes developmental timing of embryo segmentation

“…First, we built PSM-specific enhancer-mediated gene-regulatory network (eGRN) using the GRaNIE (Gene Regulatory Network Inference including Enhancers) method [29], which constructs eGRN based on co-variation of chromatin [i.e., transcription factor (TF) binding site] accessibility, TF expression and corresponding target gene expression across samples. We generated paired transcriptome [i.e., RNA sequencing (RNA-seq)] and chromatin accessibility [i.e., assay for transposase-accessible chromatin with sequence (ATAC-seq)] data from wild-type, non-cultured PSM tissues.…”

“…Enhancer-mediated gene regulatory network(eGRN) was constructed from the matched RNA-seq and ATAC-seq data (24 samples for each) ofthe PSM explants from E10.5 wild-type embryosusing the developer’s version of the now published GRaNIE package (https://bioconductor.org/packages/release/bioc/html/GRaNIE.html)[29]. Raw gene counts from RNA-seq datawere produced with a summarizeOverlaps func-tion from the GenomicAlignments R package(https://bioconductor.org/packages/release/bioc/html/GenomicAlignments.html) [62], corrected for different experimental batches usingCombat-seq function from the R package sva[63] and log2 normalised.…”

“…ATAC-seq peakcounts were generated using DiffBind R package(https://bioconductor.org/packages/DiffBind/),and peak positions were identified using MACS2software (https://genomebiology.biomedcentral.com/articles/10.1186/gb-2008-9-9-r137) [64]. Thedetails of the GRaNIE approach are describedhere [29]. Briefly, in the first step the expressionof each TF was correlated with accessibility of each of the accessible regions (=ATAC-seq peak)with and without a known binding site of theTF (foreground and background, respectively).Known binding sites were defined using theHOCOMOCO database v.10 [65].…”

DegCre: Probabilistic association of differential gene expression with regulatory regions

“…Analysts often place a threshold on TSS to CRE distances under which there is putatively high confidence for the association. These thresholds range from ∼1 kilobase (kb) to 100s of kb, with little justification provided for any given choice (Kamal et al 2023; You et al 2021; McDaniel et al 2016; Wang et al 2015). It is unlikely that any single TSS to gene distance threshold is appropriate in all contexts, and categorical thresholds generally result in limitations that impact all downstream analyses.…”

Single-cell and spatial multiomic inference of gene regulatory networks using SCRIPro

“…Recent advancements in single-cell multi-omics data have led to the development of new tools for predicting TR activity. For example, tools such as FigR[14], GRaNIE[15], DIRECT-NET[16] and GLUE[17] utilize paired or integrated multiome data as inputs and employ linear/non-linear regression methods to construct gene regulatory networks. Other tools, like CellOracle[18], offer pre-built GRNs or the ability to create custom-defined GRNs using scATAC-seq data.…”