(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org. Licensed under Creative Commons Attribution (CC BY) license. url:https://journals.plos.org/plosone/s/licenses-and-copyright ------------ McComedy: A user-friendly tool for next-generation individual-based modeling of microbial consumer-resource systems ['André Bogdanowski', 'Osnabrück University', 'Department Of Ecology', 'School Of Biology Chemistry', 'Osnabrück', 'Helmholtz-Centre For Environmental Research', 'Ufz', 'Department Of Ecological Modelling', 'Leipzig', 'Karin Frank'] Date: 2022-02 Individual-based modeling is widely applied to investigate the ecological mechanisms driving microbial community dynamics. In such models, the population or community dynamics emerge from the behavior and interplay of individual entities, which are simulated according to a predefined set of rules. If the rules that govern the behavior of individuals are based on generic and mechanistically sound principles, the models are referred to as next-generation individual-based models. These models perform particularly well in recapitulating actual ecological dynamics. However, implementation of such models is time-consuming and requires proficiency in programming or in using specific software, which likely hinders a broader application of this powerful method. Here we present McComedy, a modeling tool designed to facilitate the development of next-generation individual-based models of microbial consumer-resource systems. This tool allows flexibly combining pre-implemented building blocks that represent physical and biological processes. The ability of McComedy to capture the essential dynamics of microbial consumer-resource systems is demonstrated by reproducing and furthermore adding to the results of two distinct studies from the literature. With this article, we provide a versatile tool for developing next-generation individual-based models that can foster understanding of microbial ecology in both research and education. Microorganisms such as bacteria and fungi can be found in virtually any natural environment. To better understand the ecology of these microorganisms–which is important for several research fields including medicine, biotechnology, and conservation biology–researchers often use computer models to simulate and predict the behavior of microbial communities. Commonly, a particular technique called individual-based modeling is used to generate structurally realistic models of these communities by explicitly simulating each individual microorganism. Here we developed a tool called McComedy that helps researchers applying individual-based modeling efficiently without having to program low-level processes, thus allowing them to focus on their actual research questions. To test whether McComedy is not only convenient to use but also generates meaningful models, we used it to reproduce previously reported findings of two other research groups. Given that our results could well recapitulate and furthermore extend the original findings, we are confident that McComedy is a powerful and versatile tool that can help to address fundamental questions in microbial ecology. Funding: AB, LKM, CK, and KF are funded by the International Research School EvoCell of Osnabrück University. CK is funded by the German Research Foundation (DFG: SFB 944, P19, KO 3909/2-1, KO 3909/4-1). TB and KF are funded by the Helmholtz Research Program Terrestrial Environments. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Copyright: © 2022 Bogdanowski et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Here we present the modeling tool McComedy ( M i c robial Co mmunities, Me tabolism, and Dy namics), which constitutes a framework for individual-based modeling of microbial consumer-resource systems. A central idea of this framework is to provide generic submodels based on biological and physical principles, which we refer to as process modules and which can be combined and parametrized in a user-friendly graphical interface, resulting in ready-to-use next-generation IBMs. We tested the validity of our approach by using McComedy to implement two specific IBMs corresponding to two different studies of spatial and evolutionary dynamics of microbial communities, which involved both experiments and IBMs. For both cases, we demonstrate that the respective model constructed with McComedy was able to robustly reproduce the general results and capture the essential mechanisms underlying the microbial community dynamics in the original studies. Furthermore, we demonstrate how McComedy can be used for tackling open research questions by extending the two original studies with additional insights. IBMs can be distinguished between traditional ones and so-called next-generation IBMs [ 27 ]. Traditional IBMs are designed and parametrized on the basis of site-specific measurements (e.g. the interaction of two species is modeled according to their co-occurrence in the modeled ecosystem). This makes these models non-generic and non-transferable to other environments [ 27 ]. Next-generation IBMs overcome this drawback by constructing the individuals’ behavior from generic submodels that are based on well-understood principles from physics, chemistry, physiology, and evolutionary biology [ 24 , 27 ]. This mechanistic approach increases the propensity of the models to capture the organization and functioning of the real system (i.e. structural realism) rather than only matching empirically observed patterns [ 24 ]. In microbial ecology, this is reflected in several IBMs (e.g. [ 19 , 28 – 30 ]), which result in strikingly realistic model behavior and a thorough understanding of ecological mechanisms. Besides providing specific insights in their respective fields of application, these models demonstrate the general potential of next-generation IBMs for microbial ecology. However, building and using next-generation IBMs usually requires good knowledge in programming or proficiency with specific software tools, which hinders a more widespread application by microbial ecologists. An easy-to-use framework that facilitates the development of such IBMs from pre-implemented, tested, generic and mechanistically sound submodels could therefore contribute significantly to the field. The relatively well-understood nature of individual microorganisms in terms of movement, metabolism, and reproduction (as opposed to the more complex dynamics at the level of populations and communities) makes microbial systems particularly well-suited for simulation in IBMs. For this reason, IBMs are frequently applied to analyze different aspects of microbial ecology and evolution [ 20 ]. The simulation model results can be analyzed on different levels of organization, ranging from below (e.g. metabolic networks within individual microorganisms [ 19 , 25 , 26 ]), at (e.g. movement trajectories of individuals) and above the level of individuals (e.g. spatial distributions of entire populations or community compositions [ 16 – 18 ]). In addition, the resulting data can be directly compared to results derived from experiments, thus making IBMs very powerful to link empirical observations with theory. IBMs are commonly used to investigate the dynamics of populations or communities by simulating individual entities, which in ecology usually represent individual organisms [ 21 , 22 ]. The dynamics of populations and communities then emerge from the simulated behavior of these individuals. This bottom-up approach has been shown to be particularly useful for modeling complex systems, where individuals exhibit trait variation, adaptive behavior, or localized interactions [ 23 , 24 ]. Microbial community dynamics usually involve metabolic interactions such as the exchange of and competition for resources [ 9 , 10 ]. Focusing on those interactions, microbial communities together with the resources can be viewed as consumer-resource systems. Traditionally, consumer-resource systems are modeled using differential equations for the densities of consumer and resource species at the level of populations [ 11 , 12 ]. Such population-level equations are still applied in microbial ecology [ 13 – 15 ], but recent research of microbial consumer-resource systems is increasingly concerned with the dynamics within populations, particularly when a spatial component needs to be explicitly considered [ 16 – 19 ]. Such spatially explicit approaches can provide insight on how localized processes (e.g. cross-feeding in a structured environment [ 16 , 18 ]) shape the community on a larger scale. For that, individual-based models (IBMs) are widely applied [ 20 ]. Microbial communities are pervasive across all ecosystems and most often essential for their functioning [ 1 , 2 ]. However, the vast taxonomic diversity of their members, manifold interactions within communities and between microorganisms and their environments, as well as heterogeneities (e.g. in composition and functioning) across spatial and temporal scales pose a major challenge to understand their ecology [ 1 , 2 , 3 – 5 ]. On the other hand, a better sense of how microbial communities assemble and respond to environmental conditions is essential to fuel advance in various research fields such as medicine, biotechnology, and climate change research [ 6 – 8 ]. Results McComedy McComedy is an open-source modeling tool for developing and using IBMs of microbial communities, with a focus on consumer-resource interactions and their implications for the functioning of the corresponding communities. This tool was developed to facilitate fast and user-friendly operation as well as to grant high flexibility in model design (Fig 1). The software can be downloaded from https://git.ufz.de/bogdanow/mccomedy, where also the source code and a tutorial on how to get started are provided. To create a new IBM, the user can select several process modules, which implement biological and physical processes of relevance for microbial consumer-resource systems such as consumption or production of resources, resource diffusion through the environment, and growth of individual microorganisms. Next, parameter values of the selected process modules can be defined according to the specifics of the modeled system on the basis of empirical observation or literature. Subsequently, simulations are executed and spatially explicit data on the modeled system at discrete time points is generated. PPT PowerPoint slide PNG larger image TIFF original image Download: Fig 1. Intended workflow when using McComedy. The modeler designs an individual-based model (IBM) by selecting process modules under consideration of the research question and the current understanding of the system. The parameter values that are necessary for the simulation of the selected processes are set by the modeler, e.g. according to experimental data or literature. The resulting IBM generates spatiotemporally explicit data of the modeled microbial system. https://doi.org/10.1371/journal.pcbi.1009777.g001 A specific IBM created with McComedy describes a three-dimensional environment in which individual microorganisms (in McComedy referred to as microbes) and resources interact. Microbes are modeled as individual objects with spherical shapes and continuous positions in the environment. Specific types can be defined that differ in certain traits, such as their metabolic requirement or growth parameters. Resources are not modeled as individual particles but instead as concentrations in each grid cell of a three-dimensional grid covering the simulated environment. Resources are different metabolites that can be consumed or produced by the microbes. If necessary, the three-dimensional environment can be reduced to represent two dimensions by constraining the third dimension to just one layer of grid cells. Over the simulated time span, microbes and resources are subject to the modeled processes. These processes are encapsulated in so-called process modules, which mediate direct and indirect interactions among microbes and between microbes and the resources. Each process module simulates one component of the system dynamics, such as microbial growth or resource diffusion. In order to facilitate a flexible yet functional model design, each process module is implemented based on generic principles, which means that no ad-hoc assumptions are made for particular model applications. Instead, the dynamics of each modeled microbial system emerge entirely from the same pool of generic principles. For example, the process module Growth transforms consumed resources into biomass under consideration of a yield to be defined (cf. McComedy ODD protocol (S1 Text), 7.2.6 Growth). The module Replication divides a microbe individual into two once a critical biomass has been reached (cf. McComedy ODD protocol (S1 Text), 7.2.11 Replication). These processes are mechanistically valid regardless of the specific modeled system and are therefore preferable to alternatives, such as imposed rules or ad-hoc assumptions (e.g. a microbe replicating by chance when it is close to resources). We use the term generic principles (instead of first principles, which is also common in the literature [24]), because we do not claim that our processes are completely described by scientific laws, as we also use reasonable simplifications if we consider them mechanistically sound. McComedy does not allow for imposing higher-level processes (e.g. spatial pattern formation or density-dependent regulation of population size) as such dynamics are supposed to emerge from the generic process modules. The graphical user interface of McComedy supports a fast and user-friendly model development. The user is guided through different development stages, starting with the selection of process modules. According to the selection, McComedy shows tables containing the required parameters with editable default values. The user can also specify lists of values for single parameters and McComedy will run simulations for every combination of these parameter values. Moreover, the user can control technical settings such as the number of replicates and the configuration of the model output. The model output is generated separately for each individual simulation, in order to facilitate comparative analyses with regard to parameter variations as well as variance analyses due to stochasticity. For each simulation, the state of each microbe and resource grid cell is written into result files at predefined time intervals. The aforementioned state includes spatial coordinates, biomass, microbial type, resource concentration, as well as other properties, which allow not only for a highly-resolved and spatially explicit model analysis, but also for a direct statistical comparison with a variety of empirical data (e.g. growth kinetics, spatial patterns, functional responses, etc.). The computation time for a simulation depends mostly on the size of the simulated environment, the number of microbes included, the time step lengths of the process modules, the termination condition, and the hardware used. Simulating a microbial community for 10 virtual hours on a regular computer can take between few minutes and several days. We provide an estimate of reference computation times on a current standard computer for different representative parameter choices in S2 File. For further details on the implementation and use of McComedy please consult the Methods section as well as the ODD protocol (S1 Text). To demonstrate that the IBMs built with McComedy can capture and serve to analyze the dynamics of specific microbial systems, McComedy was used to reproduce the outcomes of two exemplary studies of microbial systems. The studies were chosen from the literature based on the close correspondence of their research questions to McComedy’s intended field of application. Thus, both studies assess spatial structuring in microbial communities as a consequence of consumer-resource interactions. We compared the outcomes of McComedy with the empirical and modeling results of the original studies. In the following, we show how McComedy can help to analyze and compare the respective results and how it can provide additional insight into the underlying mechanisms. [END] [1] Url: https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1009777 (C) Plos One. "Accelerating the publication of peer-reviewed science." Licensed under Creative Commons Attribution (CC BY 4.0) URL: https://creativecommons.org/licenses/by/4.0/ via Magical.Fish Gopher News Feeds: gopher://magical.fish/1/feeds/news/plosone/