Page contents

Published research that used Smoldyn simulations

2023


Maheshwari2

Engineering tRNA abundances for synthetic cellular systems

Maheshwari, A.J., J. Calles, S.K. Waterton, D. Endy Nature Comm. 14:4594, 2023

The authors used a modified version of Smoldyn to predict how tRNA abundances impact protein synthesis rates. Using rational design and direct RNA synthesis, they then made 21 synthetic tRNA surrogates from scratch and used a computer aided design framework to engineer translation systems that were predicted to work faster or slower. Wet lab tests of the synthetic systems showed qualitative agreement with predictions. This shows that first principles modeling combined with experiments can be used for rational design in synthetic biology.
Recouvreux

Establishment of Wnt ligand-receptor organization and cell polarity in the C. elegans embryo

Recouvreux, P., P. Pai, R. Torro, M. Ludányi, P. Mélénec, M. Boughzala, V. Bertrand, P. Lenne BioRxiv doi: https://doi.org/10.1101/2023.01.17.524363, 2023

Tissue patterning in embryos typically relies on diffusing proteins, called morphogens, that are produced in one place, transported elsewhere, and then detected. Here, the authors study how Wnt ligands, a family of signaling proteins, dynamically organize to establish cell polarity in a developing C. elegans embryonic tissue. They used quantitative live imaging to show that Wnt ligands diffuse extracellularly through the embryo over a timescale shorter than the cell cycle. Simulations supported this finding and also showed how Wnt signaling can influence anterior versus posterior patterning. This work shows how fast diffusion in the embryo can polarize target cells and supports diffusion-based long-range Wnt signaling.
Maheshwari

Colloidal Physics Modeling Reveals How Per-Ribosome Productivity Increases with Growth Rate in Escherichia coli

Maheshwari, A.J., A.M. Sunol, E. Gonzalez, D. Endy, R.N. Zia mBio 14:e02865-22, 2023

As cells are fed more nutrients so that they grow faster, they clearly need to produce proteins at a faster rate. However, it has been shown that they don't have a corresponding increase in the number of ribosomes, which implies that each ribosome must produce proteins faster. These authors explored how this works using particle-based simulations with "Colloidal Smoldyn", which is their extension of the Smoldyn software. In agreement with prior results, they found that faster growth rates lead to greater macromolecular crowding within cells and that this affects reaction rates in two opposing ways: rates increase due to higher effective reactant concentrations, and decrease due to inhibited diffusion. They showed that the former effect dominates in this case, thereby answering the question of how individual ribosomes are able to produce proteins faster with higher cell growth rates.
Abdali

Phospho-signaling couples polar asymmetry and proteolysis within a membraneless microdomain in C. crescentus

Ahmed, Y.M. and G.R. Bowman BioRxiv doi: https://doi.org/10.1101/2023.08.19.553945, 2023

Many bacterial species, including the C. crescentus bacterium investigated here, exhibit asymmetric cell division, leading to the question of how the cell directs specific regulatory proteins to the correct cell poles. Curiously, both poles contain a membraneless phase-separated microdomain, established by the polar assembly hub PopZ, through most of the cell cycle, yet many of the proteins that depend on it exhibit unipolar and transient localization. Based on many experimental observations, the authors hypothesized that the activity of the kinase/phosphatase protein CckA acts as a switch that regulates PopZ's interaction several other proteins, and that this switch is controlled by a protein that localizes to only one of the two cell poles. This hypothesis was confirmed by experiement and particle-based modeling.
Abdali

Particle-based simulations reveal two positive feedback loops allow relocation and stabilization of the polarity site during yeast mating

Guan, K., D.J. Lew, and T.C. Elston BioRxiv doi: https://doi.org/10.1101/2023.05.01.538889, 2023

This paper addresses the question of how cells are able to grow in correct directions with sufficient versatility to reorient as needed, but also sufficient positive feedback to provide strong polarization. The authors studied yeast cells, which orient their polarity sites up pheromone gradients in the course of mating. A cell's initial polarity site is often oriented incorrectly, leading to a complicated reorientation process that remains poorly understood. Particle-based simulations of the core polarity circuit revealed that molecular-level fluctuations are insufficient for reorientation on their own but that inclusion of a second pathway yielded results that agreed with experiments. This second pathway forms a positive feedback loop involving the recruitment of receptors to the cell membrane, couples polarity establishment to gradient sensing, and also allows cells to stabilize their final polarity site.
Abdali

Frequency-Domain Model of Microfluidic Molecular Communication Channels with Graphene BioFET-based Receivers

Abdali, A., and M. Kuscu arXiv 2307.04229, 2023

Molecular Communication (MC) is a an engineered communication technology that is inspired by biological systems in that diffusing molecules are used for information transfer. Using microfluidics and electronic sensors (graphene field effect transistor biosensors), the authors designed a practical MC system. Models provided insights into the dispersion and distortion of received signals, thus potentially informing the design of new frequency-domain MC techniques. The model accuracy was verified through particle-based spatial stochastic simulations.
Opalko

Simulation-based Reconstructed Diffusion unveils the effect of aging on protein diffusion in Escherichia coli

Mantovanelli, L., D.S. Linnik, M. Punter, H.J. Kojakhmetov, W.M. Smigiel, and B. Poolman BioRxiv https://doi.org/10.1101/2023.04.10.536329, 2023

This paper describes a technique called Simulation-based Reconstructed Diffusion (SbRD), which determines diffusion coefficients from microscope data by correcting for confinement effects and for the bias introduced by two-dimensional models describing a three-dimensional motion. They validated the method on simulated diffusion data. With SbRD, and a new cell detection method, the authors inferred the diffusion coefficients of several native proteins in Escherichia coli, observing slower diffusion at the cell poles, and particularly at old poles. They hypothesize that these effects are caused by aggregated or damaged proteins.
Opalko

Design principles of Cdr2 node patterns in fission yeast cells

Opalko, H., S. Geng, A.R. Hall, D. Vavylonis, and J.B. Moseley Mol. Biol. of the Cell mbc-E23, 2023

The authors investigated the localization of mitotic signaling proteins and the cytokinetic ring in fission yeast cells. They studied the positioning of nodes, which are membrane-bound multiprotein complexes and is important for timely cell cycle progression and positioning of the cytokinetic ring. They combined experimental and modeling approaches to find that Cdr2 nodes accumulate near the nucleus, and Cdr2 undergoes nucleocytoplasmic shuttling when cortical anchoring is reduced. Their model was supported by particle-based simulations based on tip inhibition, nuclear positioning, and cortical anchoring.
Yonemura

Active Transport by Cytoplasmic Dynein Maintains the Localization of MAP2 in Developing Neurons

Yonemura, Y., Y. Sakai, R. Nakata, A. Hagita-Tatsumoto, T. Miyasaka, and H. Misonou BioRxiv doi: https://doi.org/10.1101/2023.04.26.538370, 2023

MAP2 is widely used as a neuronal dendrite marker but how its localization is established and maintained has been unclear. The authors investigated how MAP2 is retained in the somatodendritic region using live cell imaging with GFP-tagged MAP2. They found migration of MAP2 toward distal dendrites via transport that depended on a serine-proline rich protein region. Inhibiting dynein showed that cytoplasmic dynein is also involved in MAP2 transport, found to occur through dynein complex binding to MAP2. Modeling based on experimental data confirmed that an intermittent active transport mechanism is essential (this modeling used the Smoldyn version that runs within VCell). The authors conclude that cytoplasmic dynein recruits and transports free MAP2 toward distal dendrites.

2022


Biosimulators

BioSimulators: a central registry of simulation engines and services for recommending specific tools

Shaikh, B., L.P. Smith, D. Vasilescu, G. Marupilla, M. Wilson, ..., and J. Karr Nucleic Acids Research 50: W108, 2022

Computational biology has struggled for many years with the wide variety of simulation tools, each with its own format, algorithms, and interface. This paper described Biosimulators, which is a central registry of simulation tools, with consistent Python, command-line, and containerized interfaces to each version of each tool. BioSimulators is based on standards such as CellML, SBML, SED-ML, and the COMBINE archive format. It supports BioNetGen, COPASI, NEURON, Tellurium, VCell, Smoldyn, and other simulators. See the Biosimulators website.
Cytoplasm

Protein diffusion in Escherichia coli cytoplasm scales with the mass of the complexes and is location dependent

Smigiel, W.M., L. Mantovanelli, D.S. Linnik, M. Punter, J. Silberberg, L. Xiang, K. Xu, and B. Poolman Science Advances 8: eabo5387, 2022

This paper investigates the structure of the E. coli cytoplasm by experimentally tracking single proteins using single-molecule displacement mapping. In this technique, the proteins are fluorescently labeled and stroboscopically illuminated to create a list of displacements from one flash to the next. Results showed that diffusion is slower near the cell poles and that it scales with the mass of the probes, but that it is not affected substantially by native protein abundance. This suggests that the cytoplasm has subdomains with varying perceived viscosities. The authors interpreted their experimental data using Smoldyn simulations.
Progesterone pathway

Stochastic model of ERK-mediated progesterone receptor translocation, clustering and transcriptional activity

Marquez-Lago, T.T. and S. Steinberg Scientific Reports 12: 11791, 2022

Progesterone receptors play key roles in the differentiation of the uterine endometrium, making progestin useful for treating endometrial cancer. To better understand the roles of progesterone receptors in these contexts, this work presents a spatial stochastic model to study the effects of progesterone receptor phosphorylation on the levels of active transcription factor. These simulations confirm prior in vitro experiments, identifying clustering as a central activating mechanism. This work was mostly performed in ChemCell, a simulator that is conceptually similar to Smoldyn but has not been updated for many years, as a tribute to Alex Slepoy, who was one of its authors. The simulation results were then confirmed with Smoldyn, which has been validated much more thoroughly.
Simularium

The Simularium Viewer: an interactive online tool for sharing spatiotemporal biological models

Lyons, B., E. Isaac, N.H. Choi, T.P. Do, J. Domingus, J. Iwasa, A. Leonard, M. Riel-Mehan, E. Rodgers, L. Schaefbauer, D. Toloudis, O. Waltner, L. Wilhelm, and G.T. Johnson Nature Methods 19: 511, 2022

This paper introduces the Simularium viewer, which is a user-friendly open source application for sharing and interrogating three-dimensional visualizations of biological simulation trajectories in a web browser. It's a platform for sharing simulation outputs with an easy-to-use interface. This is especially important for spatial modeling because these models tend to be complicated and, typically, not compatible with SBML or other standards due to the wide variety of modeling frameworks that are used. The Simularium website is compatible with models developed in Cytosim, PhsiCell, Smoldyn, SpringSaLaD, and other simulators, and includes examples of each.
Weighted ensemble

Simulation of receptor triggering by kinetic segregation shows role of oligomers and close contacts

R. Taylor, J. Allard, and E.L. Read Biophysical J. 121:1660, 2022

T cell signaling occurs when a close contact is formed between an antigen presenting cell and another cell, and this close contact can only happen once the nearby CD45 molecules have diffused away from the region around the T cell receptor. These authors studied the rate at which signaling can occur by simulating CD45 diffusion in a "region of interest" (ROI) surrounding the T cell receptor. These simulations would be extraordinarily slow normally, so the authors applied weighted ensemble averaging methods, shown in the diagram, to accelerate the simulations by many orders of magnitude. This method distributes computational effort equally among the rare events (few molecules in the ROI) and the common events (many molecules in the ROI).
Feedback

Close agreement between deterministic vs. stochastic modeling of first-passage time to vesicle fusion

Matveev, V. Biophys. J. 121:4569-4584, 2022

Calcium-dependent cell processes are inherently stochastic due to large fluctuations in calcium channel gating and other processes, but prior studies showed surprisingly close agreement between stochastic and deterministic simulations. This paper investigates this result, showing that the close agreement arises from small correlations between fluctuations of reactant molecule numbers, despite large fluctuation amplitudes. In the case investigated here, diffusion and buffering effectively decorrelated the calcium sensor and calcium concentration fluctuations.
Pipeline

Systematic comparison of modeling fidelity levels and parameter inference settings applied to negative feedback gene regulation

Coulier, A., P. Singh, M. Sturrock, and A. Hellander PLoS Comp. Biol. 18:e1010683, 2022

A persistent challenge in computational biology is in knowing what level of simulation detail is required to adequately model a problem. More detail typically produces more accurate results, but at a large computational cost, so it's generally better to use less detail when possible. These tradeoffs are particularly challenging for model parameter estimation, which is already computationally intensive. This paper develops a computational pipeline (see figure) that systematically evaluates model inference accuracy for a wide range of true known parameters and then uses it to explore inference settings for negative feedback gene regulation.

2021


Colorful cells

Exploratory polarization facilitates mating partner selection in Saccharomyces cerevisiae

Clark-Cotton, M.R., N.T. Henderson, M. Pablo, D. Ghose, T.C. Elston, and D.J. Lew Molecular Biology of the Cell 32:1048, 2021

Budding yeast cells use extracellular pheromone gradients to locate potential mating partners, which they then grow toward. This work investigates how cells polarize, such as the fact that they often polarize in unproductive directions initially, but then relocate their polarity sites until two partners' polarity sites align. Polarity development is especially complicated in dense clusters of cells, which is also the situation that is most common in nature. This work is primarily experimental (the figure shows cells of opposite mating types in green and purple, and an outlined fused zygote) but also used Smoldyn simulations, which suggested that focal secretion at polarity sites is the central step in triggering commitment with the partner cell.
Cell polarity

A novel stochastic simulation approach enables exploration of mechanisms for regulating polarity site movement

Ramirez, S.A., M. Pablo, S. Burk, D.J. Lew, T.C. Elston PLoS Comp. Biol. 17: e1008525, 2021

Many cells direct their movement or growth toward external cues, such as budding yeast cells growing toward potential mating partners in response to pheromone gradients. Within the cell, this directed growth is controlled by polarity factors that assemble into clusters at the cell membrane. These clusters are highly dynamic, with assembly, disassembly, and motion, before forming a stable polarity site that is directed toward the pheromone source. These authors investigated this using their own computationally efficient RDME (reaction-diffusion master equation) software, and validated their results using Smoldyn. Most simulations focused on a planar 2D membrane surface, but some also used a spherical surface, as shown in the figure.
Blue spheres

High-Throughput Measurements of Intra-Cellular and Secreted Cytokine from Single Spheroids Using Anchored Microfluidic Droplets

Saint-Sardos, A., S. Sart, K. Lippera, E. Brient-Litzler, S. Michelin, G. Amselem, C.N. Baroud Small 16: 2002303, 2021

This paper describes a new experimental approach for measuring the secretions from individual spheroids of human stem cells that are cultured within microfluidic droplets. It enables high-throughput quantification by using a secondary droplet that brings functionalized micro-beads into proximity with each spheroid. Models, performed with Smoldyn, investigated the molecular accumulation within the droplets and showed that physical confinement is crucial for these measurements. The model also showed that the time to achieve a measurement scales with droplet volume.
Feedback

A multiscale compartment-based model of stochastic gene regulatory networks using hitting-time analysis

Coulier, A., S. Hellander, and A. Hellander J. Chem. Phys. 154:184105, 2021

This paper describes a new simulation method for accelerating spatial simulations. It presents a multiscale model where a compartment-based model approximates a detailed spatial stochastic model. The compartment model is constructed via a first-exit time analysis on the spatial model, for which the authors used Smoldyn. The end result has computational cost that is close to that of a stochasic well-mixed model. This new method is demonstrated using the negative feedback loop for gene regulation that is shown in the diagram, showing substantial speed improvements.
T cell signaling

Trapping or slowing the diffusion of T cell receptors at close contacts initiates T cell signaling

K.Y. Chen, E. Jenkins, M. Körbel, A. Ponjavic, A.H. Lippert, A.M. Santos, N. Ashman, C. O'Brien-Ball, J. McBride, D. Klenerman, and S.J. Davis Proc. Natl. Adac. Sci. USA 118:e2024250118, 2021

The authors used live cell imaging to investigate how T cell receptor (TCR) mobility affects T cell signaling. They found that reduced mobility increases signaling. This suggests that trapping TCRs in tight cell-cell interaction spaces might initiate T cell signaling, which extends the prior understanding that signaling was tuned by TCR-ligand affinity. They used Smoldyn simulations, shown in the figure, to investigate TCR diffusion for TCRs that were bound to different adducts, and then determining fractions of cells triggered from those diffusion results.
neurons

Regional Neurodegeneration in vitro: The Protective Role of Neural Activity

Mott, R.E., C.R. von Reyn, B.L. Firestein, and D.F. Meaney Frontiers in Comp. Neuroscience 15:580107, 2021

This paper investigates the role of neural activity on recovery after neural injury. It's mostly an experimental paper, in which the authors applied microtrauma to in vitro neural tissues and imaged the results with high speed calcium imaging. They found that neurons that were more active prior to injury were more resistant to trauma and more likely to recover. They used Smoldyn simulations to investigate changes in structure and calcium dynamics that would occur with random injury, using a computational neural model that this group developed previously (Singh et al., 2011).
crowding

Quantifying the roles of space and stochasticity in computer simulations for cell biology and cellular biochemistry

Johnson, M.E., A. Chen, J.R. Faeder, P. Henning, I.I. Moraru, M. Meier-Schellersheim, R.F. Murphy, T. Prustel, J.A. Theriot, A.M. Uhrmacher Molecular Biology of the Cell 32:186, 2021

The authors review spatial stochastic simulation and then quantitatively compare several simulators using biological problems of current interest. They used Virtual Cell for non-spatial and deterministic simulations, and NERDSS, Smoldyn, MCell, and eGFRD for spatial stochastic simulations. All simulators worked well on most simple test cases (however, Smoldyn was in error for 2D reaction rates, which are beyond the scope of the simulator at present). Other tests investigated macromolecular crowding, as shown in the figure, the E. coli Min system, yeast polarization, and other systems. This paper presents an excellent overview of where the field is currently and where it needs to go.

2020


pSpatiocyte

pSpatiocyte: a high-performance simulator for intracellular reaction-diffusion systems

Arjunan, S.N.V., A. Miyauchi, K. Iwamoto, and K. Takahashi BMC Bioinformatics 21:33, 2020

Spatiocyte is a spatial stochastic simulation tool that uses a fine lattice, in which each lattice site is occupied by up to one molecule at a time. These molecules can diffuse between sites or react with molecules in neighboring sites. This paper describes a parallelized version of the software which is much faster than the original serial version (which was already quite fast). The authors compared their results against Smoldyn, finding that they got comparable results, thus verifying algorithm accuracy, and that their runtimes were 45 to 55 times faster than Smoldyn. This comparison used an 8 core computer, of which pSpatiocyte is optimized to use all 8 cores, whereas Smoldyn only runs on 1 core.
2-step method

Accurate particle-based reaction algorithms for fixed timestep simulators

Johnston, S.T., C.N. Angstmann, S.N.V. Arjunan, C.H.L. Beentjes, A. Coulier, S.A. Isaacson, A.A. Khan, K. Lipkow, and S.S. Andrews 2018 MATRIX Annals 149-164, 2020

Smoldyn's standard bimolecular reaction algorithm is fast and leads to correct reaction rates, but is inaccurate on very small size scales, while GFRD is more accurate but at the cost of high computational demands, begging the question of whether new methods could be developed that are both accurate and fast. This paper develops two such methods, both of which yield exactly correct radial distribution functions. However, they are still sufficiently challenging to implement that these authors did not implement the methods in code, and nor are these methods currently available in Smoldyn.
Bispherical absorption

Effects of surfaces and macromolecular crowding on bimolecular reaction rates

Andrews, S.S. Phys. Biol. 17:045001, 2020

It is well known that macromolecular crowding affects the rates of bimolecular reactions through the effects that volume exclusion accelerates reactions and diffusion inhibition slows reactions. This paper focuses on a third effect, which is that crowders reduce access to reactants in their immediate vicinity. It shows this effect first with a single reactant at a fixed distance from an impermeable surface, shown in the figure, and then and then with a system crowded with immobile crowders. Accounting for this effect improves agreement between analytical theory and simulations but is still somewhat in error, which prompted the author to explore other models for macromolecular crowding as well. Those behaved differently, but had comparable size errors.

2019


Yeast polarization

Ratiometric GPCR signaling enables directional sensing in yeast

Henderson, N.T., M. Pablo, D. Ghose, M.R. Clark-Cotton, T.R. Zyla, J. Nolen, T.C. Elston, D.J. Lew PLoS Biol. 17:e3000484, 2019

Yeast cells sense extracellular concentration gradients of pheromone molecules by determining which side of the cell binds more pheromone molecules. This is challenging because yeast cells are small, there are few enough receptors to create stochastic effects, and the receptors are not distributed evenly over the cell surface, but yeast nevertheless accomplish it reliably. By combining experiments, such as shown in the figure, with spatial simulations, the authors found that yeast cells detect the fraction of occupied receptors rather than absolute numbers.
NH4Cl

The counterdiffusion of HCl and NH3: An experimental and modeling analysis of topochemistry, diffusion, reaction, and phase transitions

Thompson, S., P.D. Shipman, S.P. Shipman, and T.J. Zurlinden J. Chem. Phys. 150:154306, 2019

Vapor-phase ammonium chloride and hydrochloric acid react to form solid ammonium chloride. This paper focuses on the patterns that ammonium chloride makes when the gasses diffuse together from separated sources, such as some of those shown in the figure. The authors explored these patterns experimentally, analytically, and with Smoldyn simulations. They found that the patterns arose from a moving reaction front, combined with homogeneous (gas phase) and heterogeneous (on surfaces) nucleation.

2018


MLM

Reaction-diffusion kinetics on lattice at the microscopic scale

Chew, W., K. Kazunari, M. Watabe, S.V. Muniandy, K. Takahashi, and S.N.V. Arjunan Phys. Rev. E 98:032418, 2018

Several different methods have been developed for running spatial simulations. These include continuous- space and discrete-time, such as Smoldyn, continuous-space and continuous-time, such as EGFRD, and either macroscopic or microscopic lattice-based methods. Microscopic lattice methods, in which each site is about the size of one molecule, can be very fast but have had the drawback of not producing quantitatively correct output. This paper addresses those issues by developing accurate microscopic lattice methods. The authors tested their simulation results by comparing them with results from EGFRD and Smoldyn.
Histogram

How Important Is Protein Diffusion in Prokaryotes?

Schavemaker, P.E. A.J. Boersma and B. Poolman, Frontiers in Molecular Biosciences 5:93, 2018

Although this paper doesn't directly answer the question posed in the title, it is nevertheless an excellent introduction to the topics of intracellular diffusion and diffusion-limited reactions. It is also a thorough review of experimental work in the fields. Relatively few chemical reactions in cells have been shown to be diffusion-limited, but it appears that diffusion does limit overall cell growth, such as by limiting protein elongation rates during protein synthesis. The authors used Smoldyn to estimate the amount of time that a protein takes to diffuse within a cell before finding its target.
Microscopy

Scanning Fluorescence Correlation Spectroscopy for Quantification of the Dynamics and Interactions in Tube Organelles of Living Cells

Unsay, J.D., F. Murad, E. Hermann, J. Ries, and A.J. Garcıa-Saez, ChemPhysChem 19:3273, 2018

This paper introduces a new method for quantifying the diffusion and complex formation of fluorescently labeled molecules in mitrochondrial compartments. It builds on two-focus scanning fluorescence correlation spectroscopy (SFCS), in which fluorescently labeled molecules are observed in two different microscope foci and the time correlation functions of the fluorescence in those two foci are used to compute diffusion and interaction rates. This work extends SFCS to use in tubular structures, such as mitochondria. The authors used Smoldyn simulations to check their results and eliminate possible candidate models.
FRAP

Steric exclusion and protein conformation determine the localization of plasma membrane transporters

Bianchi, F., L. Syga, G. Moiset, D. Spakman, P.E. Schavemaker, C.M. Punter, A. Seinen, A.M. van Oijen, A. Robinson, and B. Poolman, Nature Comm. 9:501, 2018

Eukaryotic plasma membranes are complex structures, including eisosomes, which are recently discovered immobile multi-protein complexes that mark the sites of endocytosis. These authors investigated protein diffusion into and out of eisosomes with single protein microscopy. Their work suggests that the distinct localization patterns found for several membrane proteins in S. cerevisiae arises from a combination of slow lateral diffusion, steric exclusion, and conditional trapping in membrane compartments. They used Smoldyn simulations to model their FRAP experiments.
Spot

Reactions, diffusion, and volume exclusion in a conserved system of interacting particles

D.B. Wilson, H. Byrne, and M. Bruna, Phys. Rev. E 97:062137, 2018

This paper explores some of the effects of macromolecular crowding on diffusion, cellular chemotaxis, and bimolecular reactions. The mathematical theories derived here agreed well with Smoldyn simulations. The simulations used Smoldyn's excluded volume capabilities. In the figure shown at the left, molecules diffused outward from an initial Gaussian profile, with different diffusion behaviors depending on whether the molecules were treated with excluded volume or not.
Wrinkle

Cell death as a trigger for morphogenesis

Aguilar, B., A. Ghaffarizadeh, C.D. Johnson, G.J. Podgorski, I. Shmulevich, and N.S. Flann, PLoS ONE 13:e0191089, 2018

Biofilm wrinkling is an interesting topic that may help build understanding about morphogenesis during development. This work examines the hypothesis that the material properties of a biofilm both power and control wrinkle formation within biofilms in response to localized cell death. This research did not use the Smoldyn simulator but used the multi-cell Biocellion simulator instead. However, the authors used the SmolCrowd utility to initially place their cells.
CamKII

Biophysical attributes that affect CaMKII activation deduced with a novel spatial stochastic simulation approach

Li, X. and W.R. Holmes, PLoS Comp. Biol. 14:e1005946, 2018

This study focuses on a calcium signaling pathway that is thought to be important for learning and memory, which involves calcium, calmodulin, and the effector protein CaMKII. The authors modified Smoldyn so that it could explicitly represent molecular binding and modification states, which enabled simple simulations of the many different states of CamKII. Their experiments showed that CaMKII activity is sensitive to calcium signal frequency and their models demonstrated how this frequency dependence relies on the amount of calcium input, calmodulin availability, and the calcium diffusion rate.

2017


Min system

Robust Min-system oscillation in the presence of internal photosynthetic membranes in cyanobacteria

MacCready, J.S., J. Schossau, K.W. Osteryoung, and D.C. Ducat, Mol. Microbiol. 103:483-503, 2017

The E. coli Min system, in which the MinC, MinD, and MinE proteins oscillate from pole to pole, has become a model system for understanding emergent protein self-organization. These authors performed the first study of Min oscillations in a different species, the cyanobacterium Synechococcus elongatus. They found that the internal photosynthetic membranes in this species did not inhibit Min protein oscillations and also suggested that Min protein affinity is greater for the plasma membrane.
dwMRI

Monte Carlo simulations of diffusion weighted MRI in myocardium: validation and sensitivity analysis

Bates, J., I. Teh, D. McClymont, P. Kohl, J.E. Schneider, and V. Grau, IEEE Transactions on Medical Imaging 36:1316-1325, 2017

Diffusion weighted magnetic resonance imaging (dwMRI) is a non-invasive technique for imaging with contrast based on the diffusion of water molecules. These authors developed a model of dwMRI and validated it by comparing simulations that were based on the model again experimental data from measurements of rat heart cells. The found good agreement, based upon similar patterns in the eigenvalues of the diffusion tensor, the mean diffusivity, and the fractional anisotropy.
Wilson image

Reactions, diffusion and volume exclusion in a heterogeneous system of interacting particles

Wilson, D.B., H. Byrne, and M. Bruna, ArXiv 1705.00004, 2017

This paper investigates the effects of macromolecular crowding on molecular diffusion and reactions. It works towards the development of macroscopic models that can be computed efficiently using partial differential equations but that are able to accurately capture the non-linear diffusion that occurs at microscopic size scales. The figure shows that simuation data (points) agreed well with their theory (lines). This work is highly mathematical.
Hair cells

In vivo ribbon mobility and turnover of Ribeye at zebrafish hair cell synapses

Graydon, C.W., U. Manor, and K.S. Kindt, Scientific Reports 7:7467, 2017

Some auditory and visual sensory cells include structures called ribbons, which are composed primarily of a protein called Ribeye. These authors studied diffusion of Ribeye within ribbons, focusing on hair cells of zebrafish lateral lines (which are similar to hair cells in ears). They measured diffusion using fluorescence recovery after photobleaching (FRAP) and then interpreted the results using Smoldyn simulations.

2016


Rate comparison

Numerical Approach to Spatial Deterministic-Stochastic Models Arising in Cell Biology

Schaff, J.C. , F. Gao, Y. Li, I.L. Novak, and B.M. Slepchenko, PLoS Comp. Biol. 12:e1005236, 2016

This paper describes the integration of the deterministic VCell simulator with the stochastic Smoldyn simulator, both of which are spatial, to create the VCell hybrid solver. The resulting simulator uses the same fixed time steps for both portions, and uses an overlapping space approach, in which the solvers address different molecular species that are within the same physical region. The hybrid solver is validated here with simple tests, and also illustrated with applications to calcium sparks, stochastically gated reactions, and spontaneous cell polarization.
Particle track

Receptor dimer stabilization by hierarchical plasma membrane microcompartments regulates cytokine signaling

You, C., T.T. Marquez-Lago, C.P. Richter, S. Wilmes, I. Moraga, K.C. Garcia, A. Leier, and J. Piehler, Science Advances 2:e1600452, 2016

Recent results have suggested that cell plasma membranes may be compartmentalized into small corrals or microcompartments by the underlying membrane cytoskeleton. This paper combined experimental and simulation approaches to investigate the consequences of those microcompartments on receptor dimerization. The authors showed that it stabilized receptor dimers and caused dissociated receptors to reassociate rapidly, thus helping maintain signaling complexes.
Hybrid method

Discrete-continuous reaction-diffusion model with mobile point-like sources and sinks

Kondrat, S., O. Zimmermann, W. Wiechert, and E. von Lieres, Eur. Phys. J. E 39:11, 2016

In many soft and biological physics applications, there are multiple distinct time and length scales. For example, enzymes are large, move slowly, and are few, while metabolites are small, fast, and abundant. The authors addressed this with a new hybrid simulation method, which is stochastic-deterministic and discrete-continuous. They demonstrate it by modelling enzyme-catalysed reactions with discrete enzymes and continuous metabolites (see figure). They validated their method by comparing simulation results from their new method against those from Smoldyn.
Crowding

Excluded volume effects in on- and off-lattice reaction-diffusion models

Meinecke, L. and M. Eriksson, IET Systems Biology 11:55-64 2016

Simulations with off-lattice Brownian dynamics (e.g. Smoldyn) are computationally expensive in crowded environments. This paper investigates the extent to which on-lattice simulations can simulate reactions and diffusion in the presence of crowders. The authors show that diffusion is slowed down in the off-lattice model since randomly distributed obstacles effectively exclude more volume than those ordered on an artificial grid. Crowded reaction rates can be both increased and decreased by the grid structure. Grid artifacts increase with increasing crowder density. The authors conclude that the computationally more efficient on-lattice simulations are accurate only for low crowder densities.
Crowding

Multiscale modeling of diffusion in a crowded environment

Meinecke, L., ArXiv 2016

This paper presents a multiscale approach to modeling diffusion and reactions rates in a crowded environment. The method combines jumping according to local first exit times and jumping on a coarser Cartesian grid. Excluded volume is modeled by a diffusion equation with space-dependent diffusion coefficients. Simulations showed that crowding molecule shape and size play a crucial role in the effects on diffusive motion. They also showed that molecular crowding can enhance or inhibit chemical reactions depending on local obstacle density fluctuations.

2015


Crowding reactions

Simulating macromolecular crowding with particle and lattice-based methods

Andrews, S.S., S.N.V. Arjunan, G. Balbo, A.T. Bittig, J. Feret, K. Kaizu, and F. Liu, In D. Gilbert, M. Heiner, K. Takahashi, and A.M. Uhrmacher, eds. Multiscale Spatial Computational Systems Biology 170-187, 2015

The authors investigated the effect of macromolecular crowding on reaction rates using 5 different simulators, both to develop a better understanding of the topic and to compare the simulators. The eGFRD algorithm was presumably very accurate but ran very slowly, Smoldyn and NL-space worked well although Smoldyn appeared to produce more accurate results, Spatiocyte was very fast but results had lattice artifacts, and KaSim worked, which was remarkable because it is typically a non-spatial method. The Smoldyn results (figure) agreed with qualitative expectations but showed that the theory that the authors investigated was incorrect.
Diffusion MRI

Sensitivity Analysis of Diffusion Tensor MRI in Simulated Rat Myocardium

Bates, J., I. Teh, P. Kohl, J.E. Schneider, and V. Grau, Lecture Notes In Computer Science 9126:120-128, 2015

Diffusion MRI (magnetic resonance imaging) is a non-invasive experimental method for visualizing diffusion in tissues, in this case in rat heart tissue. Here, the authors modeled this method using Smoldyn and found close correspondance between model and experimental results. Next, they used the simulations to test the method sensitivity. They found that the diffusivity has the greatest effect and the cross-sectional area and aspect ratio of cells are important, but the cell length and volume fraction of cells had no marked effect.
Bax model

Bax monomers form dimer units in the membrane that further self-assemble into multiple oligomeric species

Subburaj, Y., K. Cosentino, M. Axmann, E. Pedrueza-Villalmanzo, E. Hermann, S. Bleicken, J. Spatz, and A.J. García-Sáez, Nature Comm. 6:8042, 2015

Bax is a key regulator of apoptosis, mediating cytochrome c release to the cytosol via oligomerization in the outer mitochondrial membrane. These authors investigated the molecular mechanism of Bax assembly and regulation by other Bcl-2 members. They found that Bax binds to the membrane in a monomeric state and then rapidly self-assembles. They also show that other proteins, cBid and Bcl-xL, help drive Bax activity. Based on these experimental results, they developed a theoretical model, using Smoldyn, which presents a new mechanism for the molecular pathway of Bax assembly to form the apoptotic pore.
T cell motility

Linking morphodynamics and directional persistence of T lymphocyte migration

Xiaji Liu, Erik S. Welf, and Jason M. Haugh, J. R. Soc. Interface 12:20141412, 2015

These researchers experimentally investigated T lymphocyte motility, which is similar to ameboid motility. They found that cells typically turn by first creating a bifurcation of the lamellipodium at the cell's leading edge and then following one of the new projections. They proposed a model of the major interactions in this process. Simulation with Smoldyn (shown in the figure) agreed with their experimental results. The figure shows the cell rear in beige and bifurcated actin regions in green.
bistabiligy

Bistability: Requirements on Cell-Volume, Protein Diffusion, and Thermodynamics

Robert G. Endres, PLoS ONE 10:e0121681, 2015

Many cell systems, including cell cycle ones and some signaling ones, are bistable. This has been studied most often using deterministic and/or non-spatial models. This paper used Smoldyn models to investigate the more physiologically relevent case, which is for stochastic chemical reactions in the confined 3-dimensional volume of a cell, considering both bacteria and eukaryotic cells. The author finds that bistability is fragile in these conditions, often requiring finely tuned parameters, small volumes, and fast diffusion coefficients. Switching can occur upon cell growth.

2014


Explosive sensor

High Level Modeling and Simulation of a Sensor System for Vapor Trace Detection of Explosives

Drago Strle, Sensors 2014 IEEE 1320-1323, 2014

This work describes a model of a sensor system for vapor trace detection of molecules in the air using an array of functionalised capacitive sensors and an electronic detection system. The author simulated vapor diffusion, and adsorption and desorption to and from the sensors, using Smoldyn. The resulting model makes it possible to study the interactions, selectivity, and sensitivity of the sensor system in an efficient way, simplifying the design of sensor system modules for vapor trace detection of different molecules in the air.
DNA target finding

An Integrated Model of Transcription Factor Diffusion Shows the Importance of Intersegmental Transfer and Quaternary Protein Structure for Target Site Finding

Hugo G. Schmidt, Sven Sewitz, Steven S. Andrews, and Karen Lipkow, PLoS ONE 9:e108575, 2014

Transcription factors and other DNA binding proteins find their DNA binding sites faster than simple 3D diffusion allows. These authors explored this acceleration with Smoldyn models. As in prior work, they found that non-specific DNA binding followed by 1D sliding reduces finding times. They also found that intersegmental transfer, in which a transcription factor that is non-specifically bound to DNA simultaneously binds a separate DNA loop and then releases the first binding site, is also effective. Finally, they found that DNA binding proteins are enriched in dimers and tetramers, perhaps because this favors intersegmental transfer.
chromosome domains

Nucleolar tethering mediates pairing between the IgH and Myc loci

Daniel E. Strongin, Mark Groudine, Joan C. Ritland Politz, Nucleus 5:474-481, 2014

The genome is spatially organized within the eukaryotic nucleus. One aspect of this is that gene loci on different chromosomes can preferentially colocalize. Using mouse strains that have different gene arragements on their chromosomes, these authors investigated the driving force behind the colocalization of the IgH and Myc loci. They found that it arose when both loci were on chromosomes that had nucleolar organizer regions (NORs). These are sites of ribosomal DNA repeat sequences, which nucleate nucleoli. Together with simulations, these results implied that chromosome tethering to nucleoli can help colocalize genes. The figure shows one chromosome in yellow, another in blue, and nucleoli in red, in a strain where colocalization occurs.
receptor stochasticity

Mosaic physiology from developmental noise: within-organism physiological diversity as an alternative to phenotypic plasticity and phenotypic flexibility

H. Arthur Woods, J. Experiment. Biol. 217:35-45, 2014

In this paper, the author proposes that stochastic variation during organismal development causes physiological diversity within a single individual. He calls this mosaic physiology. This article reviews known mechanisms by which stochastic effects arise in, and are controlled by, biological systems. It then goes on to show how then can give rise to mosaic physiology. This provides a set of diversified phenotypes within single organisms, which may help the organism to cope with novel environmental challenges. The figure shows stochastic effects in receptor-ligand binding, a simple example of biological stochasticity.
L. lactis diffusion

Impact of osmotic stress on protein diffusion in Lactococcus lactis

Jacek T. Mika, Paul E. Schavemaker, Victor Krasnikov, Bert Poolman, Mol. Microbiol. 94:857-870, 2014

Proteins diffuse many times slower inside cells than they do in aqueous solutions due to macromolecular crowding, non-specific binding, and hydrodynamic effects. This has been investigated most thoroughly in E. coli, a Gram-negative bacterium. The authors investigated diffusion in L. lactis, a Gram-positive bacterium here. They found similar diffusion coefficients and similar cell-to-cell variation in unstressed cells, but a smaller increase upon osmotic challenge (as expected due to a higher turgor pressure). They measured intracellular diffusion with FRAP methods, which they validated with Smoldyn simulations (figure).
Lambda switch

Stochastic cellular fate decision making by multiple infecting lambda phage

Matthew L. Robb and Vahid Shahrezaei, PLOS One 9:e103636, 2014

Bacteriophage lambda is a classic system for studying cellular decision making. Upon infection, lambda can adopt the lytic state in which it reproduces rapidly in the bacterium causing cell death and the release of virus, or it can adopt the lysogenic state in which it incorporates itself into the cell genome; later, it may stochastically return to the lytic state. This decision is made through a genetic switch, shown in the figure. These authors extended prior work by investigating the role of the cell volume and bacterial growth rate on the decision. They also investigated spatial effects, which arise primarily from the slow diffusion of mRNA across a bacterium (about 10 minutes), finding that spatial effects were minimal.
Diffusion barrier

The long and viscous road: uncovering nuclear diffusion barriers in closed mitosis

Eder Zavala and Tatiana T. Marquez-Lago, PLoS Comp. Biol. 10:e1003725, 2014

Yeast cells retain their nuclear membranes during cell division, in a processs called closed mitosis. Membrane-bound proteins segregate aysmmetrically in the process, with some getting localized in the mother cell and others in the bud (dots in the figure). These authors explored mechanisms by which yeast cells might prevent protein diffusion across the division plane, and hence maintain the localization. They found that a combination of protein rings and sphingolipid domains is necessary during early anaphase, but that sphingolipid domains alone are adequate during late anaphase (figure), due to the elongated nuclear neck.

2013


Min system

Oscillations of Min-proteins in micropatterned environments: a three-dimensional particle-based stochastic simulation approach

Max Hoffmann and Ulrich S. Schwarz, Soft Matter 10:2388, 2014

One way in which E. coli bacteria find their mid-planes, so that cell division yields two equal size daughter cells, is with spatiotemporal oscillations of the Min proteins. These proteins oscillate from pole to pole, with minimal occupancy of the mid-plane. The simplicity and remarkable dynamics of this system has made it popular for spatial stochastic simulations. These authors investigated Min system operation in artificial micropatterned environments and in mutant filamentous cells, such as the one shown in the figure. This work highlights the robustness and variability of Min system oscillations, puts limits on the effect of putative division sites, and provides a computational framework for future studies.
simulation

Input-output relations in biological systems: measurement, information and the Hill equation

Steven A. Frank, Biology Direct 8:31, 2013

This paper discusses relations among Michaelis-Menten kinetics, the Hill equation, and biological information processing. A central question that the paper probes regards why Michaelis-Menten kinetics leads to linear input-output relations for low signal levels, but the Hill equation exhibits logarithmic sensitivity at these levels. This paper has numerous misconceptions and non-standard terminology but nevertheless presents some intriguing points. A particularly interesting result is that dose-response curve for the Michaelis-Menten reaction, which normally exhibits Hill equation behavior with Hill coefficient of 1, becomes ultrasensitive when reactions are substantially diffusion influenced, as shown in the figure.
TEM image

Dynamic transition states of ErbB1 phosphorylation predicted by spatial stochastic modeling

Meghan McCabe Pryor, Shalini T. Low-Nam, Ádám M. Halász, Diane S. Lidke, Bridget S. Wilson, and Jeremy S. Edwards, Biophys. J. 105:1533-1543, 2013

ErbB1 (epidermal growth factor receptor) is an important receptor for growth and development, and its overexpression can cause cancer. These authors built on their prior erbB1 single molecule tracking experiments to build a spatial stochastic model for erbB1 diffusion, dimerization, kinase activation, and phosphorylation. The figure shows erbB1 dimerization. The model yields new insight into the activation states of individual erbB1 monomers. For the most part, the authors used their own reimplementation of Smoldyn's algorithms, although they also used the Smoldyn for part of the research as well.
pore simulation

Membrane and actin reorganization in electropulse-induced cell fusion

Günther Gerisch, Mary Ecke, Ralph Neujahr, Jana Prassler, Andreas Stengl, Max Hoffmann, Ulrich S. Schwarz and Eberhard Neumann, J. Cell Science 127:4507-4517, 2013

Electric pulses induce Dictyostelium discoideum cells to fuse. The authors combined electron microscopy, fluorescence microscopy, and simulation to study the fusion pores and actin localization that arise in the membrane during cell fusion. They found that the plasma membranes of the contiguous cells become tangles of highly bent and interdigitated membranes. By imaging GFP diffusion from one cell to its neighbor, and then modeling this diffusion with Smoldyn simulations as shown in the figure, they found that membranes persist in a fusogenic state for up to 24 seconds before pores of about 3 nm are formed.

2012


crowding

Anomalous diffusion and multifractional Brownian motion: simulating molecular crowding and physical obstacles in systems biology

T.T. Marquez-Lago, A. Leier, and K. Burrage, IET Syst. Biol. 6:134-142, 2012

The study compares both fractional Brownian motion and continuous time random walks and highlights how well they can represent different types of spatial crowding and physical obstacles. Although diffusion around immovable obstacles could be reasonably characterised by a single Hurst exponent, the authors found that diffusion in a crowded environment seemed to exhibit multifractional properties in the form of a different short- and long-time behaviors. The figure shows a 2-dimensional crowded environment, developed with the Smoldyn utility program SmolCrowd.
yeast cells

Nuclear envelope morphology constrains diffusion and promotes asymmetric protein segregation in closed mitosis

Barbara Boettcher, Tatiana T. Marquez-Lago, Mathias Bayer, Eric L. Weiss, and Yves Barral, J. Cell Biol. 197:921-937, 2012

The authors investigated the establishment and maintenance of asymmetric cell division in budding yeast. Here, unlike in mammalian cells, the nuclear envelope is maintained during cell division, and divides at about the same time as the cell membrane. Using photobleaching and Smoldyn simulations, they found that diffusion barriers compartmentalize the nuclear membranes, whereas protein diffusion was well explained by the dumbbell shape of the anaphase nucleus. The figure shows a dividing cell with the nucleoplasmic GFP in green and mCherry-Tub1 spindles in red.
mushroom spine

Spatiotemporal maps of CaMKII in dendritic spines

Khan, S., T.S. Reese, N. Rajpoot, and A. Shabbir, J. Comput. Neurosci. 33:123-139, 2012

The authors investigated the sequestration of calcium calmodulin dependent kinase (CaMKII) in neural dendritic spines, which is a key cellular mechanism for the formation and storage of memories. The figure shows confocal microscopy data of GFP-CamKII localization shortly the GFP was photo-activated at the asterisk; bright green regions, such as at the spine tip, indicate CamKII sequestration, while red regions indicate low CamKII concentrations. Smoldyn simulations showed that a major cause of sequestion is the high number of cytoskeletal binding sites at spine tips, rather than high binding affinities.

2011


rate comparison

Correction factors for boundary diffusion in reaction-diffusion master equations

Leier, A. and T.T. Marquez-Lago, J. Chem. Phys. 135:134109, 2011

The two primary methods for simulating chemical reactions and diffusion are continuous-space single particle methods, like Smoldyn uses, and lattice-based reaction-diffusion master equation (RDME) methods. The latter are typically more efficient but can suffer from several artifacts. In this work, the authors presented an accurate RDME method for simulating reactive boundary conditions, which they validated using Smoldyn. The figure shows simulation results from their methods in black, analytical results in red, and Smoldyn results in gray.
low concentration high concentration

A density-dependent switch drives stochastic clustering and polarization of signaling molecules

Jilkine, A., S.B. Angenent, L.F. Wu, and S.J. Altschuler, PLoS Comput. Biol. 7:e1002271, 2011

Many cell types, including budding yeast, mammalian neutrophils, and amoeba, can spontaeously polarize in the absence of spatial cues. The authors propose that this polarization arises from positive feedback; it robustly maintains an off state at low concentrations of signaling molecules, and then switches to highly localized signaling clusters at higher concentrations. The figures show that in their model, increasing signaling molecule concentrations (right panel vs. left panel) leads to clusters of molecules that are in their active states (red).
neural synapse

Computational investigation of the changing patterns of subtype specific NMDA receptor activation during physiological glutametergic neurotransmission

Singh, P., A.J. Hockenberry, V. Tiruvadi, and D.F. Meaney, PLoS Comput. Biol. 7:e1002106, 2011

Neural NMDA receptors mediate many physiological functions, including the molecular basis for learning and memory. These receptors exist in various subtypes, which enables them to discriminate between different types of signals. Using Smoldyn models, the authors found that different receptors have different dynamic ranges, that specific subtypes dominate in long-term depression and long-term potentiation situations, and that the content of a specific subtype enhances response magnitude and fidelity during long-term potentiation. The yellow and red portions of the figure show pre- and post-synaptic regions, while dots represent receptors.
neural spine

Sequestration of CaMKII in dendritic spines in silico

Khan, S., Y. Zou, A. Amjad, A. Gardezi, C.L. Smith, C. Winters, T.S. Reese, J. Comput. Neurosci. 31:581-594, 2011

This work investigated the sequestration of calcium dependent kinase II (CamKII) in neural dendritic spines following synaptic stimulation. It looked at the effects of spine geometry on CamKII diffusion, binding of CamKII to the post-synaptic density, to the cytoskeleton, and to other CamKII proteins. Simulation results compared favorably with microscopy experiments. This work showed that self-aggregation of CamKII could provide a switch that amplifies CamKII sequestration and regulates its activity.
Ultrasensitivity

Ultrasensitivity in Multisite Phosphorylation of Membrane-Anchored Proteins

Dushek, O., P.A. van der Merwe, and V. Shahrezaei, Biophys. J. 100:1189-1197, 2011

The authors showed that multiple phosphorylation sites on a single protein can produce switch-like responses when reactions between proteins are limited by diffusion. This is the case for many reactions that occur on the plasma membrane. The figure shows that ultrasensitivity increases with as the number of protein phosphorylation sites increase. Smoldyn file.

2010


NeuroRD comparison

The role of type 4 phosphodiesterases in generating microdomains of cAMP: Large scale stochastic simulations

Oliveira, R.F., A. Terrin, G. Di Benedetto, R.C. Cannon, W. Koh, M. Kim, M. Zaccolo, and K.T. Blackwell, PLoS ONE 5:e11725, 2010

This work investigates mechanisms that increase the local concentration of cyclic AMP (cAMP), which is essential for normal neural functioning, including for synaptic plasticity. To do so, the authors developed a new spatial simulator called NeuroRD, which is a compartment-based simulator. It combines the spatial Gillespie method with tau-leaping and a new diffusion algorithm. The authors validated NeuroRD by comparing results against mass action theory and against Smoldyn simulations. These comparisons are shown in the figure.
Bar1 model

Detailed Simulations of Cell Biology with Smoldyn 2.1

Andrews, S.S., N.J. Addy, R. Brent, and A.P. Arkin, PLoS Comp. Biol. 6:e1000705, 2010

The authors investigated mating pheromone signaling between yeast cells. A central "receiver" cell is covered with receptors that bind mating pheromone, which is secreted by surrounding "sender" cells (in the figure, unbound receptors are blue, bound receptors are red, and pheromone is green). This work showed that the central cell can better locate strong pheromone emitting cells when the central cell secretes a pheromone-degrading protease because the protease cloud sharpens the local pheromone gradient.

2009


Particle counts

Modeling molecule exchange at membranes

Palm, M.M., M.N. Steijaert, H.M.M. ten Eikelder, and P.A.J. Hilbers Proceedings of the Third International Conference on the Foundations of Systems Biology in Engineering, 2009

This paper focuses on molecule adsorption and desorption at surfaces, comparing results from PDE type and particle-based simulations. When they began their work, Smoldyn placed desorbed molecules at the surface, which the authors showed to disagree with PDE results. Smoldyn was improved during the course of this work to place desorbed molecules at the correct distance from the surface, which the authors showed led to excellent agreement with PDEs. This paper also shows that stochastic simulations are needed when there are relatively few individual molecules.
Simulated FRAP

Introducing simulated cellular architecture to the quantitative analysis of fluorescent microscopy

DePristo, M.A., L. Chang, R.D. Vale, S.M. Khan, and K. Lipkow, Prog. Biophys. Mol. Biol. 100:25-32, 2009

The authors used simulations to analyze and quantify experimental Fluorescence Recovery After Photobleaching (FRAP) data for proteins in the E. coli chemotaxis system. They quantified protein diffusion coefficients of 2 μm2/s and assessed turnover rates between cytoplasmic proteins and membrane-associated protein clusters. The figure shows simulated CheY-YFP fluorescence in a protein cluster as it is photobleached and then recovers from protein exchange. In general, simulations can be useful tools for the quantitative analysis of fluorescent microscopy experiments.

2008


Protein gradients

Model for Protein Concentration Gradients in the Cytoplasm

Lipkow, K. and D.J. Odde, Cell. Mol. Bioeng. 1:84-92, 2008

Using simulations, the authors showed that several of the E. coli chemotaxis proteins (CheY and CheZ) likely have stable protein concentration gradients across the length of cells. This arises from protein activation that is localized to one cell pole and activation-dependent protein complexation that affects protein diffusion coefficients. These mechanisms, and similar ones, occur for many other proteins as well, so stable protein concentration gradients within cells are likely to be widespread.

2006


Hair cell PMCA2

Rapid Turnover of Stereocilia Membrane Proteins: Evidence from the Trafficking and Mobility of Plasma Membrane Ca2+-ATPase 2

Grati, M., M.E. Schneider, K. Lipkow, E.E. Strehler, R.J. Wenthold, and B. Kachar, J. Neuroscience 26:6386-6395, 2006

This work combined experimental and modeling approaches to investigate membrane protein turnover in stereocilia (fine hairs in the inner ear that convert mechanical motion to electrical signals) to better understand stereocilia recovery from trauma, such as loud noises. The authors focused on the spatial distribution, mobility, and trafficking of the PMCA2 protein, which is both abundant and essential to stereocilia function. They found that PMCA2 exhibits rapid turnover, which supports other evidence that stereocilia undergo rapid continuous renewal. The figure shows PMCA2 (green dots) distribution on an unrolled stereocilia membrane over time.
Lotka Volterra scale Lotka Volterra dynamics

Simulating cell biology

Andrews, S.S. and A.P. Arkin, Current Biology 16:R523-R527, 2006

This is a primer on how to simulate cell biology systems. It presents the basics of ODE (ordinary differential equation) modeling, stochastic modeling, spatial modeling, and spatial-stochastic modeling. Of particular interest, it presents simulations of the same Lotka-Volterra predator-prey system that used each of the different methods. The results differ dramatically. Deterministic simulations show nearly sinusoidal oscillations, non-spatial stochastic simulations show oscillations with increasing amplitudes, and particle-tracking (Smoldyn; figure show here) simulations show a characteristic boom-and-bust pattern.
CheZ localization

Changing Cellular Location of CheZ Predicted by Molecular Simulations

Lipkow, K., PLoS Comp. Biol. 2:e39, 2006

In the E. coli chemotaxis system, the CheY messenger protein is phosphorylated at a receptor cluster near one cell pole and is dephosphorylated by the CheZ protein. The author used simulations to explore the signaling behaviors that arise with different CheZ localizations. She found that the model that best agreed with experiments included CheZ proteins that oligomerize and localize to the receptor cluster upon cellular stimulation. The black dots in the figure are CheZ oligomers (also, pink dots are CheY, red are CheYp, and green are non-oligomeric CheZ). This creates a negative feedback loop which improves system precision, robustness, and adaptation.

2005


FliM occupancy

Simulated Diffusion of Phosphorylated CheY through the Cytoplasm of Escherichia coli

Lipkow, K., S.S. Andrews, and D. Bray, J. Bact. 187:45-53, 2005

The authors developed a spatial stochastic simulation of the central E. coli chemotaxis system. It includes the core proteins and reactions, a polar-localized receptor cluster, and several flagellar motors in the cell membrane. Increasing cytoplasmic crowding, due to ribosomes and other macromolecules, resulted in slower protein diffusion, steepened concentration gradients, and substantial differences in signal propagation delays to motors at different positions (figure).

Publications that describe Smoldyn software and algorithms

Singh image

Python interfaces for the Smoldyn simulator

Singh, D., and S.S. Andrews, Bioinformatics 38:291-293, 2022

This article describes a Python application programming interface (API) for Smoldyn. It improves integration with other software tools, such as Jupyter notebooks, other Python code libraries and other simulators. The API includes low-level functions that closely mimic the existing C/C++ API and higher-level functions that are more convenient to use. These latter functions follow modern object-oriented Python conventions.
Andrews image

Rule-based modeling using wildcards in the Smoldyn simulator

Andrews, S.S., Methods in Mol. Biol. 1945:179-202, 2019

This article presents a new approach to rule-based modeling, which is a method for automatically generating new chemical species for modified and complexed molecules. This approach is based on wildcards that match to species names, much as wildcards can match to file names in computer operating systems. It is simple to use and remarkably powerful. This article demonstrates the approach through examples for: signaling systems, protein complexation, polymerization, nucleic acid sequence copying and mutation, the "SMILES" chemical notation, and others. It is implemented in Smoldyn for both the generate-first and on-the-fly expansion.
Andrews image

Particle-based stochastic simulators

Andrews, S.S., Encyclopedia of Computational Neuroscience, 2018

This article defines and tells about particle-based stochastic simulators, aimed toward the neuroscience community but quite accessible to others as well. It includes a thorough comparison of the currently maintained particle-based simulators, which are MCell, Smoldyn, eGFRD, SpringSaLaD, and ReaDDy. This table was compiled from publications about these packages, discussions with their authors, and several tests of each package. A more detailed version of the table, including executable versions of all of the input files, is available on this web site, here.
Andrews image Andrews image

Smoldyn: particle-based simulation with rule-based modeling, improved molecular interaction and a library interface

Andrews, S.S., Bioinformatics 33:710, 2017

This paper describes several additions to the Smoldyn software. Smoldyn now supports rule-based modeling with a convenient wildcard method, and also with the BioNetGen package with extensions for spatial simulation. New algorithms for simulating the diffusion of surface-bound molecules and molecules with excluded volume provide excellent accuracy (figure). In addition, Smoldyn supports single-molecule tracking simulations. Finally, the Smoldyn source code can be accessed through a C/C++ language library interface.
Seeliger image

Enabling surface dependent diffusion in spatial simulations using Smoldyn

Seeliger, C. and N. Le Novère, BMC Res. Notes 8:752, 2015

The authors added support for diffusion coefficients of surface-bound molecules to depend on the surface to which the molecule is bound. For example, in the figure, certain surface-bound molecules diffuse more slowly within the blue region of the surface than in other portions of the surface. This is useful for simulating lipid rafts or similar structures. Unfortunately, this addition is not included in the current Smoldyn release, but only in the authors' modification of Smoldyn 2.22, which is otherwise largely obsolete.
Hybrid Smoldyn

Multiscale reaction-diffusion simulations with Smoldyn

Robinson, M., S.S. Andrews, and R. Erban, Bioinformatics 31:2406-2408, 2015

The authors integrated a lattice-based simulation solver with Smoldyn's existing particle-based simulations. This is an adjacent-space type of hybrid method, meaning that different regions of space get simulated with different levels of detail. They used recently devloped methods (by Flegg and Erban) to address the boundary between two different simulation regions. This accelerates simulations several-fold. It is very easy to use.
Smoldyn example

Spatial and stochastic cellular modeling with the Smoldyn simulator

Andrews, S.S., In Bacterial Molecular Networks: Methods and Protocols, van Helden at al. (eds.) Methods in Molecular Biology 804:519-542, 2012

This paper describes the use of the Smoldyn simulation program. It focuses on the basic concepts, while leaving syntax details to the Smoldyn User's Manual. This figure was taken from an example simulation that presents the use of reactions, molecule-surface interactions, compartments, and other Smoldyn features. This paper presents an especially useful Notes section that gives suggestions on choosing simulation parameters (e.g. diffusion coefficient and reaction rates) and methods for improving computational performance.
GPU Smoldyn

Smoldyn on Graphics Processing Units: Massively Parallel Brownian Dynamics Simulation

Dematté, L., IEEE Trans. Comput. Biol. Bioinform. 9:655-667, 2012

This is a more thorough presentation of the work described below by the same author. In brief, he parallelized the core Smoldyn code to run on Graphics Processing Units (GPUs), thus accelerating it 130-fold.
GPU Smoldyn

Accelerating the Smoldyn spatial stochastic biochemical reaction network simulator using GPUs

Gladkov, D.V., S. Alberts, R.M. D'Souza, and S. Andrews, Proceedings of the 19th High Performance Computing Symposia, 2011

The authors accelerated Smoldyn simulations by about 200 fold by parallelizing the core code to run on Graphics Processing Units (GPUs). This work was performed independently of Dematté's work, listed below. The figure shows code acceleration for different Smoldyn algorithms. Diffusion has the best performace, while bimolecular reactions have the worst (but still excellent) performance.
GPU Smoldyn

Parallel particle-based reaction diffusion: A GPU implementation

Dematté, L., 2010 Ninth International Workshop on Parallel and Distributed Methods in Verification/2010 Second International Workshop on High Performance Computing 67-77, 2010

The author parallelized the core Smoldyn code to run on Graphics Processing Units (GPUs). This can accelerate simulations by factors of up to 130, thus enabling the efficient simulation of models that would be impractical to run on conventional computer hardware. The figure explains efficient memory organization for GPUs.
Smoldyn accuracy

Detailed Simulations of Cell Biology with Smoldyn 2.1

Andrews, S.S., N.J. Addy, R. Brent, and A.P. Arkin, PLoS Comp. Biol. 6:e1000705, 2010

This paper (which is also listed above) describes the Smoldyn software. It gives an overview of the software algorithms, features, and performance. It shows that Smoldyn is very accurate, is more than twice as fast as either MCell or ChemCell, and has a wide range of features. This figure shows the number of molecules adsorbed to a surface over time for several adsorption coefficients; close accuracy between simulation data (dots) and theoretical calculations (lines) show Smoldyn's accuracy. The inset shows that simulation noise matches theoretical calculations.
Adsorption

Accurate particle-based simulation of adsorption, desorption and partial transmission

Andrews, S.S., Phys. Biol. 6:046015, 2009

This paper describes methods for quantitatively simulating the adsorption of molecules to surfaces, the desorption of molecules from surfaces, and the partial transmission of molecules through surfaces. Central results of the paper include look-up tables that give a molecule's adsorption or transmission probability based on the surface's adsorption or transmission coefficient, the molecule's diffusion coefficient, the simulation time step, and the rate of the reverse desorption or transmission process. Smoldyn uses these algorithms for molecule-surface interactions.
Simulator comparison

Computational methods for diffusion-influenced biochemical reactions

Dobrzynski, M., J.V. Rodríguez, J.A. Kaandorp, and J.G. Blom, Bioinformatics 23:1969-1977, 2007

The authors compared the three major spatial stochastic simulation methods, which are particle-based methods (with Smoldyn), spatial Gillespie methods (with GMP), and Green's Function Reaction Dynamics (GFRD; with custom code). They found that GFRD is most accurate and that the relative accuracy between particle-based and spatial Gillepsie methods depends on the latter's spatial resolution; with fine resolution, spatial Gillespie methods can be better than particle-based methods. For systems with many molecules, GFRD is extremely slow, while Smoldyn takes about twice as long as GMP to yield results with comparable accuracy. The authors noted that particle-based methods are particularly flexible for creating realistic geometries.
Bimolecular reaction simulation

Stochastic simulation of chemical reactions with spatial resolution and single molecule detail

Andrews, S.S. and D. Bray, Phys. Biol. 1:137, 2004

The authors derived several of the central algorithms for particle-based simulation, all designed for high accuracy and good computational efficiency. Methods are presented for simulating diffusion and chemical reactions, where the reactions can be 0th, 1st, or 2nd order and they can be irreversible or reversible. This figure shows one of the paper's central results, which is a look-up table that enables simulated bimolecular chemical reaction rates (y-axis) to be determined from simulation parameters (x-axis) for both irreversible (bottom line) and reversible reactions (other lines).