GRaNIE and GRaNPA: Inference and evaluation of enhancer-mediated gene regulatory networks applied to study macrophages

Abstract

Among the biggest challenges in the post-GWAS (genome-wide association studies) era is the interpretation of disease-associated genetic variants in non-coding genomic regions. Enhancers have emerged as key players in mediating the effect of genetic variants on complex traits and diseases. Their activity is regulated by a combination of transcription factors (TFs), epigenetic changes and genetic variants. Several approaches exist to link enhancers to their target genes, and others that infer TF-gene connections. However, we currently lack a framework that systematically integrates enhancers into TF-gene regulatory networks. Furthermore, we lack an unbiased way of assessing whether inferred regulatory interactions are biologically meaningful. Here we present two methods, implemented as user-friendly R-packages, for building and evaluating enhancer-mediated gene regulatory networks (eGRNs) called GRaNIE (Gene Regulatory Network Inference including Enhancers - https://git.embl.de/grp-zaugg/GRaNIE) and GRaNPA (Gene Regulatory Network Performance Analysis - https://git.embl.de/grp-zaugg/GRaNPA), respectively. GRaNIE jointly infers TF-enhancer, enhancer-gene and TF-gene interactions by integrating open chromatin data such as ATAC-Seq or H3K27ac with RNA-seq across a set of samples (e.g. individuals), and optionally also Hi-C data. GRaNPA is a general framework for evaluating the biological relevance of TF-gene GRNs by assessing their performance for predicting cell-type specific differential expression. We demonstrate the power of our tool-suite by investigating gene regulatory mechanisms in macrophages that underlie their response to infection, and their involvement in common genetic diseases including autoimmune diseases.Among the biggest challenges in the post-GWAS (genome-wide association studies) era is the interpretation of disease-associated genetic variants in non-coding genomic regions. Enhancers have emerged as key players in mediating the effect of genetic variants on complex traits and diseases. Their activity is regulated by a combination of transcription factors (TFs), epigenetic changes and genetic variants. Several approaches exist to link enhancers to their target genes, and others that infer TF-gene connections. However, we currently lack a framework that systematically integrates enhancers into TF-gene regulatory networks. Furthermore, we lack an unbiased way of assessing whether inferred regulatory interactions are biologically meaningful. Here we present two methods, implemented as user-friendly R-packages, for building and evaluating enhancer-mediated gene regulatory networks (eGRNs) called GRaNIE (Gene Regulatory Network Inference including Enhancers - https://git.embl.de/grp-zaugg/GRaNIE) and GRaNPA (Gene Regulatory Network Performance Analysis - https://git.embl.de/grp-zaugg/GRaNPA), respectively. GRaNIE jointly infers TF-enhancer, enhancer-gene and TF-gene interactions by integrating open chromatin data such as ATAC-Seq or H3K27ac with RNA-seq across a set of samples (e.g. individuals), and optionally also Hi-C data. GRaNPA is a general framework for evaluating the biological relevance of TF-gene GRNs by assessing their performance for predicting cell-type specific differential expression. We demonstrate the power of our tool-suite by investigating gene regulatory mechanisms in macrophages that underlie their response to infection, and their involvement in common genetic diseases including autoimmune diseases.Among the biggest challenges in the post-GWAS (genome-wide association studies) era is the interpretation of disease-associated genetic variants in non-coding genomic regions. Enhancers have emerged as key players in mediating the effect of genetic variants on complex traits and diseases. Their activity is regulated by a combination of transcription factors (TFs), epigenetic changes and genetic variants. Several approaches exist to link enhancers to their target genes, and others that infer TF-gene connections. However, we currently lack a framework that systematically integrates enhancers into TF-gene regulatory networks. Furthermore, we lack an unbiased way of assessing whether inferred regulatory interactions are biologically meaningful. Here we present two methods, implemented as user-friendly R-packages, for building and evaluating enhancer-mediated gene regulatory networks (eGRNs) called GRaNIE (Gene Regulatory Network Inference including Enhancers - https://git.embl.de/grp-zaugg/GRaNIE) and GRaNPA (Gene Regulatory Network Performance Analysis - https://git.embl.de/grp-zaugg/GRaNPA), respectively. GRaNIE jointly infers TF-enhancer, enhancer-gene and TF-gene interactions by integrating open chromatin data such as ATAC-Seq or H3K27ac with RNA-seq across a set of samples (e.g. individuals), and optionally also Hi-C data. GRaNPA is a general framework for evaluating the biological relevance of TF-gene GRNs by assessing their performance for predicting cell-type specific differential expression. We demonstrate the power of our tool-suite by investigating gene regulatory mechanisms in macrophages that underlie their response to infection, and their involvement in common genetic diseases including autoimmune diseases.

Decoding gene regulation in the mouse embryo using single-cell multi-omics

“…Second, we used the in silico ChIP-seq library to link TFs to cis-regulatory elements (ATAC peaks). Third, we linked cis-regulatory elements to potential target genes by genomic proximity (here a maximum distance of 50kb), which is a reasonable approximation in the absence of 3D chromatin contact information (Janssens et al, 2022; Kamal et al, 2021). This results in a directed network where each parent node corresponds to a TF, and each child node corresponds to a target gene.…”

SCENIC+: single-cell multiomic inference of enhancers and gene regulatory networks

“…A classical GRN contains by definition the ‘upstream’ TFs, their genomic binding sites, and their target genes 1,88 . However, only few GRNs have been characterized to the level of detail where they include cis-regulatory elements as nodes, so that the TF is linked to its target gene via its binding site in a regulatory region 3,22,42 .…”

“…We created a new benchmark dataset to compare eGRN inference methods using the ENCODE core cell lines and foresee that future methods can be easily benchmarked on the same data. We compared SCENIC+ with other methods, including CellOracle 39 , FigR 41 , GRaNIE 42 , and Pando 40 . All these methods solve certain aspects of the eGRN inference problem and focus on particular sub-problems.…”

“…FigR proposes a new concept of co-regulated chromatin hubs, called TF-DORCs (Domains of Regulatory Chromatin), thereby focusing more on chromatin interactions rather than individual TF binding sites and enhancers 41 . Pando 40 and GRaNIE 42 have more similar goals compared to SCENIC+, but GRaNIE was originally designed for bulk RNA/ ATAC data (although we have shown here that for small single-cell data sets it performs accurately). We believe that having multiple methods available to solve eGRN mapping is an advantage for the community, and since each method has its own strengths, their combination may yield further improvements.…”

“…We simulated 500 single-cell multiomics profiles by randomly sampling 50,000 reads and 20,000 fragments from each bulk RNA-seq and ATAC-seq profiles, respectively; resulting in a data set with 4,000 simulated single-cell cells. We also ran other methods that predict GRN/eGRNs using multiomics data and TF motifs as input, namely CellOracle 39 , Pando 40 , FigR 41 , and GRaNIE 42 ; and as a baseline we included an updated version of SCENIC, which uses scRNA-seq and TF motifs but no chromatin accessibility 19,20 (Fig 3a,b, Suppl Note 1). We first assessed how many TFs were recovered by each method, and how many target regions and genes were predicted per TF (Fig 3c , S5a-e).…”