A survey of gene regulatory networks modelling methods: from differential equations, to Boolean and qualitative bioinspired models

Gene Regulatory Networks (GRNs) represent the interactions among genes regulating the activation of specific cell functionalities, such as reception of (chemical) signals or reaction to environmental changes. Studying and understanding these processes is crucial: they are the fundamental mechanism at the basis of cell functioning, and many diseases are based on perturbations or malfunctioning of some gene regulation activities. In this paper, we provide an overview on computational approaches to GRN modelling and analysis. We start from the biological and quantitative modelling background notions, recalling differential equations and the Gillespie’s algorithm. Then, we describe more in depth qualitative approaches such as Boolean networks and some computer science formalisms, including Petri nets, P systems and reaction systems. Our aim is to introduce the reader to the problem of GRN modelling and to guide her/him along the path that goes from classical quantitative methods, through qualitative methods based on Boolean network, up to some of the most relevant qualitative computational methods to understand the advantages and limitations of the different approaches.

A Multi-stage Representation of Cell Proliferation as a Markov Process

The stochastic simulation algorithm commonly known as Gillespie’s algorithm (originally derived for modelling well-mixed systems of chemical reactions) is now used ubiquitously in the modelling of biological processes in which stochastic effects play an important role. In well-mixed scenarios at the sub-cellular level it is often reasonable to assume that times between successive reaction/interaction events are exponentially distributed and can be appropriately modelled as a Markov process and hence simulated by the Gillespie algorithm. However, Gillespie’s algorithm is routinely applied to model biological systems for which it was never intended. In particular, processes in which cell proliferation is important (e.g. embryonic development, cancer formation) should not be simulated naively using the Gillespie algorithm since the history-dependent nature of the cell cycle breaks the Markov process. The variance in experimentally measured cell cycle times is far less than in an exponential cell cycle time distribution with the same mean. Here we suggest a method of modelling the cell cycle that restores the memoryless property to the system and is therefore consistent with simulation via the Gillespie algorithm. By breaking the cell cycle into a number of independent exponentially distributed stages, we can restore the Markov property at the same time as more accurately approximating the appropriate cell cycle time distributions. The consequences of our revised mathematical model are explored analytically as far as possible. We demonstrate the importance of employing the correct cell cycle time distribution by recapitulating the results from two models incorporating cellular proliferation (one spatial and one non-spatial) and demonstrating that changing the cell cycle time distribution makes quantitative and qualitative differences to the outcome of the models. Our adaptation will allow modellers and experimentalists alike to appropriately represent cellular proliferation—vital to the accurate modelling of many biological processes—whilst still being able to take advantage of the power and efficiency of the popular Gillespie algorithm.

A multi-stage representation of cell proliferation as a Markov process

The stochastic simulation algorithm commonly known as Gillespie’s algorithm (originally derived for modelling well-mixed systems of chemical reactions) is now used ubiquitously in the modelling of biological processes in which stochastic effects play an important role. In well mixed scenarios at the sub-cellular level it is often reasonable to assume that times between successive reaction/interaction events are exponentially distributed and can be appropriately modelled as a Markov process and hence simulated by the Gillespie algorithm. However, Gillespie’s algorithm is routinely applied to model biological systems for which it was never intended. In particular, processes in which cell proliferation is important (e.g. embryonic development, cancer formation) should not be simulated naively using the Gillespie algorithm since the history-dependent nature of the cell cycle breaks the Markov process. The variance in experimentally measured cell cycle times is far less than in an exponential cell cycle time distribution with the same mean.Here we suggest a method of modelling the cell cycle that restores the memoryless property to the system and is therefore consistent with simulation via the Gillespie algorithm. By breaking the cell cycle into a number of independent exponentially distributed stages we can restore the Markov property at the same time as more accurately approximating the appropriate cell cycle time distributions. The consequences of our revised mathematical model are explored analytically as far as possible. We demonstrate the importance of employing the correct cell cycle time distribution by recapitulating the results from two models incorporating cellular proliferation (one spatial and one non-spatial) and demonstrating that changing the cell cycle time distribution makes quantitative and qualitative differences to the outcome of the models. Our adaptation will allow modellers and experimentalists alike to appropriately represent cellular proliferation - vital to the accurate modelling of many biological processes - whilst still being able to take advantage of the power and efficiency of the popular Gillespie algorithm.

Path-wise versus kinetic modeling for equilibrating non-Langevin jump-type processes

We discuss two independent methods of solution of a master equation whose biased jump transition rates account for long jumps of Lévy-stable type and admit a Boltzmannian (thermal) equilibrium to arise in the large time asymptotics of a probability density function ρ(x, t). Our main goal is to demonstrate a compatibility of a direct solution method (an explicit, albeit numerically assisted, integration of the master equation) with an indirect pathwise procedure, recently proposed in [Physica A 392, 3485, (2013)] as a valid tool for a dynamical analysis of non-Langevin jump-type processes. The path-wise method heavily relies on an accumulation of large sample path data, that are generated by means of a properly tailored Gillespie’s algorithm. Their statistical analysis in turn allows to infer the dynamics of ρ(x, t). However, no consistency check has been completed so far to demonstrate that both methods are fully compatible and indeed provide a solution of the same dynamical problem. Presently we remove this gap, with a focus on potential deficiencies (various cutoffs, including those upon the jump size) of approximations involved in simulation routines and solutions protocols.

A Model of the Quorum Sensing System in Vibrio fischeri Using P Systems

Quorum sensing is a cell-density-dependent gene regulation system that allows an entire population of bacterial cells to communicate in order to regulate the expression of certain or specific genes in a coordinated way depending on the size of the population. We present a model of the quorum sensing system in Vibrio fischeri using a variant of membrane systems called P systems. In this framework each bacterium and the environment are represented by membranes, and the rules are applied according to an extension of Gillespie's algorithm called the multicompartmental Gillespie's algorithm. This algorithm runs on more than one compartment and takes into account the disturbance produced when chemical substances diffuse from one compartment or region to another one. Our approach allows us to examine the individual behavior of each bacterium as an agent as well as the emergent behavior of the colony as a whole and the processes of swarming and recruitment. Our simulations show that at low cell densities bacteria remain dark, while at high cell densities some bacteria start to produce light and a recruitment process takes place that makes the whole colony of bacteria do so. Our computational modeling of quorum sensing could provide insights leading to new applications where multiple agents need to robustly and efficiently coordinate their collective behavior based only on very limited information about the local environment.