Stiffness detection and reduction in discrete stochastic simulation of biochemical systems

Typical multiscale biochemical models contain fast-scale and slow-scale reactions, where "fast" reactions fire much more frequently than "slow" ones. This feature often causes stiffness in discrete stochastic simulation methods such as Gillespie's algorithm and the Tau-Leaping method leading to inefficient simulation. This paper proposes a new strategy to automatically detect stiffness and identify species that cause stiffness for the Tau-Leaping method, as well as two stiffness reduction methods. Numerical results on a stiff decaying dimerization model and a heat shock protein regulation model demonstrate the efficiency and accuracy of the proposed methods for multiscale biochemical systems.     

A constant-time kinetic Monte Carlo algorithm for simulation of large biochemical reaction networks

The time evolution of species concentrations in biochemical reaction networks is often modeled using the stochastic simulation algorithm (SSA) [Gillespie, J. Phys. Chem. 81, 2340 (1977)]. The computational cost of the original SSA scaled linearly with the number of reactions in the network. Gibson and Bruck developed a logarithmic scaling version of the SSA which uses a priority queue or binary tree for more efficient reaction selection [Gibson and Bruck, J. Phys. Chem. A 104, 1876 (2000)]. More generally, this problem is one of dynamic discrete random variate generation which finds many uses in kinetic Monte Carlo and discrete event simulation. We present here a constant-time algorithm, whose cost is independent of the number of reactions, enabled by a slightly more complex underlying data structure. While applicable to kinetic Monte Carlo simulations in general, we describe the algorithm in the context of biochemical simulations and demonstrate its competitive performance on small- and medium-size networks, as well as its superior constant-time performance on very large networks, which are becoming necessary to represent the increasing complexity of biochemical data for pathways that mediate cell function.     

Torsion in biochemical reaction networks

This article starts in Part I with a simple example of two biochemical reaction networks that are indistinguishable at the macroscopic level but are different at the molecular level and are shown to have significantly different kinetic properties. So, if one completely ignores the fact that reactions advance in discrete steps at the molecular level, then one can fail to distinguish between networks with widely different kinetics. In part II biochemical reaction networks are treated in a general way to discover what property of a network, only seen at the molecular level, affects its kinetics. It is shown that every such network has a unique torsion group which can be described numerically and readily determined by a programmable computation. If the group is found to be the singleton {0} (as is most often the case in practice), then the network is said to be torsion-free and its kinetic properties unaffected by ignoring its discrete character. A chemical reaction network has to be represented algebraically to calculate its torsion group. If the network is to be understood only at the macroscopic level, it can be placed in the context of real vector spaces, but to recognize its discrete character and its torsion group, each vector space is replaced by a discrete subset of that space, where each molecule can be recognized as a distinct and indivisible entity. Next, the process of calculating a torsion group is shown in several cases, including the example in part I. In this particular case it is shown to have the torsion group with 2 elements, reflecting the fact that the substrate molecules become product molecules 2 at a time, with the result that the overall macroscopic reaction is R ⇔ T, whereas at the molecular level it is 2R ⇔ 2T. In general, however, the torsion group of a biochemical reaction network can be any finite additive group, which is a property of the network that can only be seen at the molecular level. Finally, this fact is demonstrated by showing how to construct a hypothetical, but plausible, biochemical reaction network that has any given finite additive group as its torsion group.     

Green's-function reaction dynamics: a particle-based approach for simulating biochemical networks in time and space

We have developed a new numerical technique, called Green's-function reaction dynamics (GFRD), that makes it possible to simulate biochemical networks at the particle level and in both time and space. In this scheme, a maximum time step is chosen such that only single particles or pairs of particles have to be considered. For these particles, the Smoluchowski equation can be solved analytically using Green's functions. The main idea of GFRD is to exploit the exact solution of the Smoluchoswki equation to set up an event-driven algorithm, which combines in one step the propagation of the particles in space with the reactions between them. The event-driven nature allows GFRD to make large jumps in time and space when the particles are far apart from each other. Here, we apply the technique to a simple model of gene expression. The simulations reveal that spatial fluctuations can be a major source of noise in biochemical networks. The calculations also show that GFRD is highly efficient. Under biologically relevant conditions, GFRD is up to five orders of magnitude faster than conventional particle-based techniques for simulating biochemical networks in time and space. GFRD is not limited to biochemical networks. It can also be applied to a large number of other reaction-diffusion problems.     

Exact results for noise power spectra in linear biochemical reaction networks

We present a simple method for determining the exact noise power spectra and related statistical properties for linear chemical reaction networks. The method is applied to reaction networks which are representative of biochemical processes such as gene expression. We find, for example, that a post-translational modification reaction can reduce the noise associated with gene expression. Our results also indicate how to coarse grain networks by the elimination of fast reactions. In this context we have discovered a breakdown of the sum rule which relates the noise power spectrum to the total noise. The breakdown can be quantified by a sum rule deficit, which is found to be universal, and can be attributed to the high-frequency noise in the fast reactions.     

Improved delay-leaping simulation algorithm for biochemical reaction systems with delays

In biochemical reaction systems dominated by delays, the simulation speed of the stochastic simulation algorithm depends on the size of the wait queue. As a result, it is important to control the size of the wait queue to improve the efficiency of the simulation. An improved accelerated delay stochastic simulation algorithm for biochemical reaction systems with delays, termed the improved delay-leaping algorithm, is proposed in this paper. The update method for the wait queue is effective in reducing the size of the queue as well as shortening the storage and access time, thereby accelerating the simulation speed. Numerical simulation on two examples indicates that this method not only obtains a more significant efficiency compared with the existing methods, but also can be widely applied in biochemical reaction systems with delays.     

Analysis of droplet behavior in a de-oiling hydrocyclone

A mechanical separation process in a de-oiling hydrocyclone is described in which disperse oil droplets are separated from a continuous water phase. This separation process is influenced by droplet breakage and coalescence. Based on experimental data and simulation results in a stirred tank, a modified breakage model, which can be applied to droplet breakage in the de-oiling hydrocyclone, is developed. Then, a simulation model is developed coupling the numerical solution of the flow field in the hydrocyclone based on computational fluid dynamics (CFD) with population balances. The homogenous discrete method and the inhomogeneous discrete method are applied for solving the population balance model (PBM). The investigations show that the numerical results obtained by the simulation model coupled with the modified PBM using the inhomogeneous discrete method are in good accordance with experimental data under a high flow rate. According to this simulation model, the effect of three different inlet designs on the separation efficiency of the de-oiling hydrocyclone has been discussed. The results indicate that the separation efficiency of the de-oiling hydrocyclone can be improved with an appropriate inlet design.     

The stochastic quasi-steady-state assumption: reducing the model but not the noise

Highly reactive species at small copy numbers play an important role in many biological reaction networks. We have described previously how these species can be removed from reaction networks using stochastic quasi-steady-state singular perturbation analysis (sQSPA). In this paper we apply sQSPA to three published biological models: the pap operon regulation, a biochemical oscillator, and an intracellular viral infection. These examples demonstrate three different potential benefits of sQSPA. First, rare state probabilities can be accurately estimated from simulation. Second, the method typically results in fewer and better scaled parameters that can be more readily estimated from experiments. Finally, the simulation time can be significantly reduced without sacrificing the accuracy of the solution.     

Oscillations in non-mass action kinetics models of biochemical reaction networks arising from pairs of subnetworks

It is well known that oscillations in models of biochemical reaction networks can arise as a result of a single negative cycle. On the other hand, methods for finding general network conditions for potential oscillations in large biochemical reaction networks containing many cycles are not well developed. A biochemical reaction network with any number of species is represented by a simple digraph and is modeled by an ordinary differential equation (ODE) system with non-mass action kinetics. The obtained graph-theoretic condition generalizes the negative cycle condition for oscillations in ODE models to the existence of a pair of subnetworks, where each subnetwork contains an even number of positive cycles. The technique is illustrated with a model of genetic regulation.     

Elastic Properties of Semi-Flexible Chains and Networks

Summary: Elastic properties of noncharged polymers of stiffness ranging from flexible to rigid chains are determined from Monte Carlo simulations. The discrete wormlike chain (WLC) model with self-interacting units is applied to chains of intermediate lengths. Elastic free energy and the force-extension profiles of chains of variable stiffness are computed in an isometric ensemble. Occurrence of a plateau on the force-extension curves at intermediate chain stiffness is noted. Qualitative differences are found between force profiles from simulations and from the standard (ideal) WLC model. The single-chain data on influence of bending stiffness were employed in the three-chain model of networks. Stress-strain relations for networks show a highly nonlinear behavior with the marked strain-stiffening effect.     

An equation-free probabilistic steady-state approximation: dynamic application to the stochastic simulation of biochemical reaction networks

Stochastic chemical kinetics more accurately describes the dynamics of "small" chemical systems, such as biological cells. Many real systems contain dynamical stiffness, which causes the exact stochastic simulation algorithm or other kinetic Monte Carlo methods to spend the majority of their time executing frequently occurring reaction events. Previous methods have successfully applied a type of probabilistic steady-state approximation by deriving an evolution equation, such as the chemical master equation, for the relaxed fast dynamics and using the solution of that equation to determine the slow dynamics. However, because the solution of the chemical master equation is limited to small, carefully selected, or linear reaction networks, an alternate equation-free method would be highly useful. We present a probabilistic steady-state approximation that separates the time scales of an arbitrary reaction network, detects the convergence of a marginal distribution to a quasi-steady-state, directly samples the underlying distribution, and uses those samples to accurately predict the state of the system, including the effects of the slow dynamics, at future times. The numerical method produces an accurate solution of both the fast and slow reaction dynamics while, for stiff systems, reducing the computational time by orders of magnitude. The developed theory makes no approximations on the shape or form of the underlying steady-state distribution and only assumes that it is ergodic. We demonstrate the accuracy and efficiency of the method using multiple interesting examples, including a highly nonlinear protein-protein interaction network. The developed theory may be applied to any type of kinetic Monte Carlo simulation to more efficiently simulate dynamically stiff systems, including existing exact, approximate, or hybrid stochastic simulation techniques.     

A perturbation analysis of rate theory of self-regulating genes and signaling networks

A thorough kinetic analysis of the rate theory for stochastic self-regulating gene networks is presented. The chemical master equation kinetic model in terms of a coupled birth-death process is deconstructed into several simpler kinetic modules. We formulate and improve upon the rate theory of self-regulating genes in terms of perturbation theory. We propose a simple five-state scheme as a faithful caricature that elucidates the full kinetics including the "resonance phenomenon" discovered by Walczak et al. [Proc. Natl. Acad. Sci. U.S.A. 102, 18926 (2005)]. The same analysis can be readily applied to other biochemical networks such as phosphorylation signaling with fluctuating kinase activity. Generalization of the present approach can be included in multiple time-scale numerical computations for large biochemical networks.     

Molecular dynamics extended for fluctuating networks: Application to water

Molecular simulation models are increasingly important tools in efforts to understand the role that water plays in biochemical processes. However, existing models of water have limited capacity to deal with the characteristics of hydrogen bond networks. This article proposes a new fluctuating network (FN) algorithm as an extension of the standard molecular dynamics algorithm. The new algorithm allows for the simulation of a molecular system based on an underlying network, such as the hydrogen bond network in water. This algorithm distinguishes strong from weak network connections, applying a potential that best describes the specific connection behavior. We model liquid water with this new technique using a single‐site, isotropic, short‐range potential. We successfully reproduce liquid water's signature molecular spacing (as represented by the radial distribution function) and characterize its dynamic properties including the exponential hydrogen bond lifetime distribution, diffusion rate, and average hydrogen bonds per molecule. The FN algorithm allows exploration of the behavior of networked systems where explicit coordination limits are required. As such it could also be used to model covalent interactions, reaction dynamics, and applied to simulation of cellular networks. © 2012 Wiley Periodicals, Inc.     

混沌系统的变量变化率反馈控制方法研究

设计了一种控制连续非线性系统中混沌的新方法--变量变化率脉冲反馈(VRPF)方法.介绍了VRPF方法的控制原理以及反馈系数和脉冲间隔的选择技巧.将此方法应用到BZ反应3D模型系统混沌的控制中,计算机仿真模拟显示,通过恰当地选择反馈系数和脉冲间隔,可以将系统稳定在1p、2p、3p、4p、…、2n×3mp (n、m为整数)这样不同的周期轨道,从而使系统的功率谱也由混沌态时的连续谱转变为具有分立单峰的分立谱.此外,仿真模拟还发现VRPF方法具有极宽的控制域.     

The sorting direct method for stochastic simulation of biochemical systems with varying reaction execution behavior

A key to advancing the understanding of molecular biology in the post-genomic age is the development of accurate predictive models for genetic regulation, protein interaction, metabolism, and other biochemical processes. To facilitate model development, simulation algorithms must provide an accurate representation of the system, while performing the simulation in a reasonable amount of time. Gillespie's stochastic simulation algorithm (SSA) accurately depicts spatially homogeneous models with small populations of chemical species and properly represents noise, but it is often abandoned when modeling larger systems because of its computational complexity. In this work, we examine the performance of different versions of the SSA when applied to several biochemical models. Through our analysis, we discover that transient changes in reaction execution frequencies, which are typical of biochemical models with gene induction and repression, can dramatically affect simulator performance. To account for these shifts, we propose a new algorithm called the sorting direct method that maintains a loosely sorted order of the reactions as the simulation executes. Our measurements show that the sorting direct method performs favorably when compared to other well-known exact stochastic simulation algorithms.     

A variational approach to the stochastic aspects of cellular signal transduction

Cellular signaling networks have evolved to cope with intrinsic fluctuations, coming from the small numbers of constituents, and the environmental noise. Stochastic chemical kinetics equations govern the way biochemical networks process noisy signals. The essential difficulty associated with the master equation approach to solving the stochastic chemical kinetics problem is the enormous number of ordinary differential equations involved. In this work, we show how to achieve tremendous reduction in the dimensionality of specific reaction cascade dynamics by solving variationally an equivalent quantum field theoretic formulation of stochastic chemical kinetics. The present formulation avoids cumbersome commutator computations in the derivation of evolution equations, making the physical significance of the variational method more transparent. We propose novel time-dependent basis functions which work well over a wide range of rate parameters. We apply the new basis functions to describe stochastic signaling in several enzymatic cascades and compare the results so obtained with those from alternative solution techniques. The variational Ansatz gives probability distributions that agree well with the exact ones, even when fluctuations are large and discreteness and nonlinearity are important. A numerical implementation of our technique is many orders of magnitude more efficient computationally compared with the traditional Monte Carlo simulation algorithms or the Langevin simulations.     

On the precision of quasi steady state assumptions in stochastic dynamics

Many biochemical networks have complex multidimensional dynamics and there is a long history of methods that have been used for dimensionality reduction for such reaction networks. Usually a deterministic mass action approach is used; however, in small volumes, there are significant fluctuations from the mean which the mass action approach cannot capture. In such cases stochastic simulation methods should be used. In this paper, we evaluate the applicability of one such dimensionality reduction method, the quasi-steady state approximation (QSSA) [L. Menten and M. Michaelis, "Die kinetik der invertinwirkung," Biochem. Z 49, 333369 (1913)] for dimensionality reduction in case of stochastic dynamics. First, the applicability of QSSA approach is evaluated for a canonical system of enzyme reactions. Application of QSSA to such a reaction system in a deterministic setting leads to Michaelis-Menten reduced kinetics which can be used to derive the equilibrium concentrations of the reaction species. In the case of stochastic simulations, however, the steady state is characterized by fluctuations around the mean equilibrium concentration. Our analysis shows that a QSSA based approach for dimensionality reduction captures well the mean of the distribution as obtained from a full dimensional simulation but fails to accurately capture the distribution around that mean. Moreover, the QSSA approximation is not unique. We have then extended the analysis to a simple bistable biochemical network model proposed to account for the stability of synaptic efficacies; the substrate of learning and memory [J. E. Lisman, "A mechanism of memory storage insensitive to molecular turnover: A bistable autophosphorylating kinase," Proc. Natl. Acad. Sci. U.S.A. 82, 3055-3057 (1985)]. Our analysis shows that a QSSA based dimensionality reduction method results in errors as big as two orders of magnitude in predicting the residence times in the two stable states.     

A comparative study of the master equation approach and the discrete galerkin method for the simulation of free radical polymerization

The present article deals with the mathematical treatment of free radical polymerization reactions. As a typical example the synthesis of poly(methyl methacrylate) under realistic experimental conditions is investigated. Since the mathematical treatment of the kinetic rate equations raises severe numerical problems, alternative approaches are required. In this paper two of these methods, i.e. the discrete Galerkin method and the master equation approach, are compared. The discrete Galerkin method circumvents difficulties encountered by the direct integration of the kinetic rate equations but requires much a priori knowledge of the chemical reaction system. Within the framework of the master equation approach the polymerization reaction is regarded as a stochastic process. For the simulation of this stochastic process a modified algorithm is presented. The example of the polymerization of methyl methacrylate shows that the master equation approach is an efficient tool in the simulation of free radical polymerization reactions.     

Reaction Brownian dynamics and the effect of spatial fluctuations on the gain of a push-pull network

Brownian Dynamics algorithms have been widely used for simulating systems in soft-condensed matter physics. In recent times, their application has been extended to the simulation of coarse-grained models of biochemical networks. In these models, components move by diffusion and interact with one another upon contact. However, when reactions are incorporated into a Brownian dynamics algorithm, care must be taken to avoid violations of the detailed-balance rule, which would introduce systematic errors in the simulation. We present a Brownian dynamics algorithm for simulating reaction-diffusion systems that rigorously obeys detailed balance for equilibrium reactions. By comparing the simulation results to exact analytical results for a bimolecular reaction, we show that the algorithm correctly reproduces both equilibrium and dynamical quantities. We apply our scheme to a "push-pull" network in which two antagonistic enzymes covalently modify a substrate. Our results highlight that spatial fluctuations of the network components can strongly reduce the gain of the response of a biochemical network.     

Threshold Sensing through a Synthetic Enzymatic Reaction–Diffusion Network

A wet stamping method to precisely control concentrations of enzymes and inhibitors in place and time inside layered gels is reported. By combining enzymatic reactions such as autocatalysis and inhibition with spatial delivery of components through soft lithographic techniques, a biochemical reaction network capable of recognizing the spatial distribution of an enzyme was constructed. The experimental method can be used to assess fundamental principles of spatiotemporal order formation in chemical reaction networks.