One model fits all: Combining inference and simulation of gene regulatory networks

Abstract

The rise of single-cell data highlights the need for a nondeterministic view of gene expression, while offering new opportunities regarding gene regulatory network inference. We recently introduced two strategies that specifically exploit time-course data, where single-cell profiling is performed after a stimulus: HARISSA, a mechanistic network model with a highly efficient simulation procedure, and CARDAMOM, a scalable inference method seen as model calibration. Here, we combine the two approaches and show that the same model driven by transcriptional bursting can be used simultaneously as an inference tool, to reconstruct biologically relevant networks, and as a simulation tool, to generate realistic transcriptional profiles emerging from gene interactions. We verify that CARDAMOM quantitatively reconstructs causal links when the data is simulated from HARISSA, and demonstrate its performance on experimental data collected on in vitro differentiating mouse embryonic stem cells. Overall, this integrated strategy largely overcomes the limitations of disconnected inference and simulation.

Modeling relaxation experiments with a mechanistic model of gene expression

“…Inference of such GRNs is a notoriously difficult task (see e.g. [29]), and performing relaxation experiments from such complex objects is yet to be done.…”

scTIGER: A Deep-Learning Method for Inferring Gene Regulatory Networks from Case versus Control scRNA-seq Datasets

“…Simulated datasets allow us to have control over the underlying regulatory relationships, noise levels, and other factors, providing a baseline for comparison and evaluation. In brief, we can define the regulatory relationships among target genes and their corresponding regulators, simulate gene expression data based on these relationships and experimental conditions using mathematical models [29,73] such as differential equations or Boolean networks, and introduce appropriate levels of noise to mimic the variability and measurement errors observed in scRNA-seq data. Additionally, we can incorporate technical variations typical in scRNA-seq experiments, such as dropout events and amplification biases, to make the simulated datasets more realistic.…”

Gene regulatory network inference from single-cell data using optimal transport

“…Despite the overly simple model class, this result gives hope that single cell data can accelerate transcriptomics research. Methods for GRN inference from single-cell data have been proposed, based on, for example, information-theoretic considerations [16], regression [31, 47], co-expression [62], and dynamical models [45, 1, 51, 9, 65]. In a benchmarking study [55], regression-based methods seemed to be the most consistent, and were the best performers in the inference task from real scRNA-Seq data.…”