Network inference using steady-state data and Goldbeter–Koshland kinetics

Abstract

Motivation: Network inference approaches are widely used to shed light on regulatory interplay between molecular players such as genes and proteins. Biochemical processes underlying networks of interest (e.g. gene regulatory or protein signalling networks) are generally nonlinear. In many settings, knowledge is available concerning relevant chemical kinetics. However, existing network inference methods for continuous data are typically rooted in convenient statistical formulations which do not exploit chemical kinetics to guide inference. Results: Here we present an approach to network inference for steady-state data that is rooted in nonlinear descriptions of biochemical mechanism. We use equilibrium analysis of chemical kinetics to obtain functional forms that are in turn used to infer networks using steady-state data. The approach we propose is directly applicable to conventional steady-state gene expression or proteomic data and does not require knowledge of either network topology or any kinetic parameters; both are simultaneously learned from data. We illustrate the approach in the context of protein phosphorylation networks, using data simulated from a recent mechanistic model and proteomic data from cancer cell lines. In the former, the true network is known and used for assessment, whilst in the latter results are compared against known biochemistry. We find that the proposed methodology is more effective at estimating network topology than methods based on linear models. Availability: MATLAB

Causal Inference by using Invariant Prediction: Identification and Confidence Intervals

“…(), Oates et al . () and Oates and Mukherjee (). For time dynamic processes the causal interpretation of a structural equation model (or a directed acyclic graph) for cross‐sectional data may be problematic, and alternatives need to be considered (Aalen et al ., , , ; Røysland, ; Sokol and Hansen, 2014).…”

Exploiting network topology for large-scale inference of nonlinear reaction models

“…Large-scale network inference with nonlinear forward models, in particular, has seen relatively limited attention [6]. A notable effort from Oates et al [17] applies Bayesian model selection to ODE models systematically generated from potential species interactions, using reversible-jump MCMC (a general across-model sampling method) to simultaneously sample network topologies and their underlying rate parameters. Their approach employs standard methods for nested models: when adding reactions, the prior distribution is used to propose new reaction rates, and when swapping reactions, previous rate parameter values are simply retained.…”