Beyond GLMs: A Generative Mixture Modeling Approach to Neural System Identification

Generalized linear models (GLMs) represent a popular choice for the probabilistic characterization of neural spike responses. While GLMs are attractive for their computational tractability, they also impose strong assumptions and thus only allow for a limited range of stimulus-response relationships to be discovered. Alternative approaches exist that make only very weak assumptions but scale poorly to high-dimensional stimulus spaces. Here we seek an approach which can gracefully interpolate between the two extremes. We extend two frequently used special cases of the GLM—a linear and a quadratic model—by assuming that the spike-triggered and non-spike-triggered distributions can be adequately represented using Gaussian mixtures. Because we derive the model from a generative perspective, its components are easy to interpret as they correspond to, for example, the spike-triggered distribution and the interspike interval distribution. The model is able to capture complex dependencies on high-dimensional stimuli with far fewer parameters than other approaches such as histogram-based methods. The added flexibility comes at the cost of a non-concave log-likelihood. We show that in practice this does not have to be an issue and the mixture-based model is able to outperform generalized linear and quadratic models.     

Promises and limits of an agency perspective in evolutionary developmental biology

An agent-based perspective in the study of complex systems is well established in diverse disciplines, yet is only beginning to be applied to evolutionary developmental biology. In this essay, we begin by defining agency and associated terminology formally. We then explore the assumptions and predictions of an agency perspective, apply these to select processes and key concept areas relevant to practitioners of evolutionary developmental biology, and consider the potential epistemic roles that an agency perspective might play in evo devo. Throughout, we discuss evidence supportive of agential dynamics in biological systems relevant to evo devo and explore where agency thinking may enrich the explanatory reach of research efforts in evolutionary developmental biology.     

Modelling the mechanisms behind the key epidemiological processes of the conifer pathogen Heterobasidion annosum

The Heterobasidion annosum species complex is a widely distributed group of fungal conifer pathogens causing root and butt rots. We studied the key processes of Heterobasidion epidemiology by compiling models that rely on biological processes. Models were included in the mechanistic model with stochastic elements, Hmodel, simulating the fungal dynamics in even-aged Norway spruce stands. The results from the modelling and stand-level simulations indicated that primary infections are affected by the stump size and spore deposition. In addition, we found that Heterobasidion dynamics at the scale of the stand are driven by several infections in large stumps, rather than by numerous infections in small stumps. We assessed the need for quantitative results in Heterobasidion biology, especially in the spread mechanisms, to support the development of complex mechanistic models.     

Sensitivity summation theorems for stochastic biochemical reaction systems

We investigate how stochastic reaction processes are affected by external perturbations. We describe an extension of the deterministic metabolic control analysis (MCA) to the stochastic regime. We introduce stochastic sensitivities for mean and covariance values of reactant concentrations and reaction fluxes and show that there exist MCA-like summation theorems among these sensitivities. The summation theorems for flux variances is shown to depend on the size of the measurement time window (?) within which reaction events are counted for measuring a single flux. It is found that the degree of the ?-dependency can become significant for processes involving multi-time-scale dynamics and is estimated by introducing a new measure of time-scale separation. This ?-dependency is shown to be closely related to the power-law scaling observed in flux fluctuations in various complex networks.     

Experimental quantification of the fluid dynamics in blood-processing devices through 4D-flow imaging: A pilot study on a real oxygenator/heat-exchanger module

The performance of blood-processing devices largely depends on the associated fluid dynamics, which hence represents a key aspect in their design and optimization. To this aim, two approaches are currently adopted: computational fluid-dynamics, which yields highly resolved three-dimensional data but relies on simplifying assumptions, and in vitro experiments, which typically involve the direct video-acquisition of the flow field and provide 2D data only. We propose a novel method that exploits space- and time-resolved magnetic resonance imaging (4D-flow) to quantify the complex 3D flow field in blood-processing devices and to overcome these limitations.We tested our method on a real device that integrates an oxygenator and a heat exchanger. A dedicated mock loop was implemented, and novel 4D-flow sequences with sub-millimetric spatial resolution and region-dependent velocity encodings were defined. Automated in house software was developed to quantify the complex 3D flow field within the different regions of the device: region-dependent flow rates, pressure drops, paths of the working fluid and wall shear stresses were computed.Our analysis highlighted the effects of fine geometrical features of the device on the local fluid-dynamics, which would be unlikely observed by current in vitro approaches. Also, the effects of non-idealities on the flow field distribution were captured, thanks to the absence of the simplifying assumptions that typically characterize numerical models.To the best of our knowledge, our approach is the first of its kind and could be extended to the analysis of a broad range of clinically relevant devices.     

Evolution of taxis responses in virtual bacteria: non-adaptive dynamics

Bacteria are able to sense and respond to a variety of external stimuli, with responses that vary from stimuli to stimuli and from species to species. The best-understood is chemotaxis in the model organism Escherichia coli, where the dynamics and the structure of the underlying pathway are well characterised. It is not clear, however, how well this detailed knowledge applies to mechanisms mediating responses to other stimuli or to pathways in other species. Furthermore, there is increasing experimental evidence that bacteria integrate responses from different stimuli to generate a coherent taxis response. We currently lack a full understanding of the different pathway structures and dynamics and how this integration is achieved. In order to explore different pathway structures and dynamics that can underlie taxis responses in bacteria, we perform a computational simulation of the evolution of taxis. This approach starts with a population of virtual bacteria that move in a virtual environment based on the dynamics of the simple biochemical pathways they harbour. As mutations lead to changes in pathway structure and dynamics, bacteria better able to localise with favourable conditions gain a selective advantage. We find that a certain dynamics evolves consistently under different model assumptions and environments. These dynamics, which we call non-adaptive dynamics, directly couple tumbling probability of the cell to increasing stimuli. Dynamics that are adaptive under a wide range of conditions, as seen in the chemotaxis pathway of E. coli, do not evolve in these evolutionary simulations. However, we find that stimulus scarcity and fluctuations during evolution results in complex pathway dynamics that result both in adaptive and non-adaptive dynamics depending on basal stimuli levels. Further analyses of evolved pathway structures show that effective taxis dynamics can be mediated with as few as two components. The non-adaptive dynamics mediating taxis responses provide an explanation for experimental observations made in mutant strains of E. coli and in wild-type Rhodobacter sphaeroides that could not be explained with standard models. We speculate that such dynamics exist in other bacteria as well and play a role linking the metabolic state of the cell and the taxis response. The simplicity of mechanisms mediating such dynamics makes them a candidate precursor of more complex taxis responses involving adaptation. This study suggests a strong link between stimulus conditions during evolution and evolved pathway dynamics. When evolution was simulated under conditions of scarce and fluctuating stimulus conditions, the evolved pathway contained features of both adaptive and non-adaptive dynamics, suggesting that these two types of dynamics can have different advantages under distinct environmental circumstances.     

How Random Is Social Behaviour? Disentangling Social Complexity through the Study of a Wild House Mouse Population

Out of all the complex phenomena displayed in the behaviour of animal groups, many are thought to be emergent properties of rather simple decisions at the individual level. Some of these phenomena may also be explained by random processes only. Here we investigate to what extent the interaction dynamics of a population of wild house mice (Mus domesticus) in their natural environment can be explained by a simple stochastic model. We first introduce the notion of perceptual landscape, a novel tool used here to describe the utilisation of space by the mouse colony based on the sampling of individuals in discrete locations. We then implement the behavioural assumptions of the perceptual landscape in a multi-agent simulation to verify their accuracy in the reproduction of observed social patterns. We find that many high-level features – with the exception of territoriality – of our behavioural dataset can be accounted for at the population level through the use of this simplified representation. Our findings underline the potential importance of random factors in the apparent complexity of the mice''s social structure. These results resonate in the general context of adaptive behaviour versus elementary environmental interactions.     

Global asymptotic stability of plant-seed bank models

Many plant populations have persistent seed banks, which consist of viable seeds that remain dormant in the soil for many years. Seed banks are important for plant population dynamics because they buffer against environmental perturbations and reduce the probability of extinction. Viability of the seeds in the seed bank can depend on the seed’s age, hence it is important to keep track of the age distribution of seeds in the seed bank. In this paper we construct a general density-dependent plant-seed bank model where the seed bank is age-structured. We consider density dependence in both seedling establishment and seed production, since previous work has highlighted that overcrowding can suppress both of these processes. Under certain assumptions on the density dependence, we prove that there is a globally stable equilibrium population vector which is independent of the initial state. We derive an analytical formula for the equilibrium population using methods from feedback control theory. We apply these results to a model for the plant species Cirsium palustre and its seed bank.     

Dynamics of Mitochondrial RNA-Binding Protein Complex in Trypanosoma brucei and Its Petite Mutant under Optimized Immobilization Conditions

There are a variety of complex metabolic processes ongoing simultaneously in the single, large mitochondrion of Trypanosoma brucei. Understanding the organellar environment and dynamics of mitochondrial proteins requires quantitative measurement in vivo. In this study, we have validated a method for immobilizing both procyclic stage (PS) and bloodstream stage (BS) T. brucei brucei with a high level of cell viability over several hours and verified its suitability for undertaking fluorescence recovery after photobleaching (FRAP), with mitochondrion-targeted yellow fluorescent protein (YFP). Next, we used this method for comparative analysis of the translational diffusion of mitochondrial RNA-binding protein 1 (MRP1) in the BS and in T. b. evansi. The latter flagellate is like petite mutant Saccharomyces cerevisiae because it lacks organelle-encoded nucleic acids. FRAP measurement of YFP-tagged MRP1 in both cell lines illuminated from a new perspective how the absence or presence of RNA affects proteins involved in mitochondrial RNA metabolism. This work represents the first attempt to examine this process in live trypanosomes.     

Many facades of CTCF unified by its coding for three-dimensional genome architecture

CCCTC-binding factor (CTCF) is a multifunctional zinc finger protein that is conserved in metazoan species. CTCF is consistently found to play an important role in many diverse biological processes. CTCF/cohesin-mediated active chromatin ‘loop extrusion’ architects three-dimensional (3D) genome folding. The 3D architectural role of CTCF underlies its multifarious functions, including developmental regulation of gene expression, protocadherin (Pcdh) promoter choice in the nervous system, immunoglobulin (Ig) and T-cell receptor (Tcr) V(D)J recombination in the immune system, homeobox (Hox) gene control during limb development, as well as many other aspects of biology. Here, we review the pleiotropic functions of CTCF from the perspective of its essential role in 3D genome architecture and topological promoter/enhancer selection. We envision the 3D genome as an enormous complex architecture, with tens of thousands of CTCF sites as connecting nodes and CTCF proteins as mysterious bonds that glue together genomic building parts with distinct articulation joints. In particular, we focus on the internal mechanisms by which CTCF controls higher order chromatin structures that manifest its many façades of physiological and pathological functions. We also discuss the dichotomic role of CTCF sites as intriguing 3D genome nodes for seemingly contradictory ‘looping bridges’ and ‘topological insulators’ to frame a beautiful magnificent house for a cell's nuclear home.     

BDI-modelling of complex intracellular dynamics

A BDI-based continuous-time modelling approach for intracellular dynamics is presented. It is shown how temporalised BDI-models make it possible to model intracellular biochemical processes as decision processes. By abstracting from some of the details of the biochemical pathways, the model achieves understanding in nearly intuitive terms, without losing veracity: classical intentional state properties such as beliefs, desires and intentions are founded in reality through precise biochemical relations. In an extensive example, the complex regulation of Escherichia coli vis-à-vis lactose, glucose and oxygen is simulated as a discrete-state, continuous-time temporal decision manager. Thus a bridge is introduced between two different scientific areas: the area of BDI-modelling and the area of intracellular dynamics.     

Application of three independent consequential LCA approaches to the agricultural sector in Luxembourg

Purpose

Consequential Life Cycle Assessment (C-LCA) is a “system modelling approach in which activities in a product system are linked so that activities are included in the product system to the extent that they are expected to change as a consequence of a change in demand”. Hence, C-LCA focuses on micro-economic actions linked to macro-economic consequences, by identifying the (marginal) suppliers and technologies prone to be affected by variable scale changes in the demand of a product. Detecting the direct and indirect environmental effects due to changes in the production system is not an easy task. Hence, researchers have combined the consequential perspective with different econometric models. Therefore, the aim of this study is to assess an increase in biocrops cultivation in Luxembourg using three different consequential modelling approaches to understand the benefits, drawbacks and assumptions linked to each approach as applied to the case study selected.

Methods

Firstly, a partial equilibrium (PE) model is used to detect changes in land cultivation based on the farmers’ revenue maximisation. Secondly, another PE model is proposed, which considers a different perspective aiming at minimising a total adaptation cost (so-called opportunity cost) to satisfy a given new demand of domestically produced biofuel. Finally, the consequential system delimitation for agricultural LCA approach, as proposed by Schmidt (Int J Life Cycle Assess 13:350–364, 2008), is applied.

Results and discussion

The two PE models present complex shifts in crop rotation land use changes (LUCs), linked to the optimisation that is performed, while the remaining approach has limited consequential impact on changes in crop patterns since the expert opinion decision tree constitutes a simplification of the ongoing LUCs. However, environmental consequences in the latter were considerably higher due to intercontinental trade assumptions recommended by the experts that were not accounted for in the economic models. Environmental variations between the different scenarios due to LUCs vary based on the different expert- or computational-based assumptions. Finally, environmental consequences as compared with the current state-of-the-art are lame due to the limited impact of the shock within the global trade market.

Conclusions

The use of several consequential modelling approaches within the same study may help widen the interpretation of the advantages or risks of applying a specific change to a production system. In fact, different models may not only be good alternatives in terms of comparability of scenarios and assumptions, but there may also be room for complementing these within a unique framework to reduce uncertainties in an integrated way.     

CONAN: A Tool to Decode Dynamical Information from Molecular Interaction Maps

The analysis of contacts is a powerful tool to understand biomolecular function in a series of contexts, from the investigation of dynamical behavior at equilibrium to the study of nonequilibrium dynamics in which the system moves between multiple states. We thus propose a tool called CONtact ANalysis (CONAN) that, from molecular dynamics (MD) trajectories, analyzes interresidue contacts, creates videos of time-resolved contact maps, and performs correlation, principal component, and cluster analysis, revealing how specific contacts relate to functionally relevant states sampled by MD. We present how CONAN can identify features describing the dynamics of ubiquitin both at equilibrium and during mechanical unfolding. Additionally, we show the analysis of MD trajectories of an α-synuclein mutant peptide that undergoes an α-β conformational transition that can be easily monitored using CONAN, which identifies the multiple states that the peptide explores along its conformational dynamics. The high versatility and ease of use of the software make CONAN a tool that can significantly facilitate the understanding of the complex dynamical behavior of proteins or other biomolecules. CONAN and its documentation are freely available for download on GitHub.     

Comparison of Deterministic and Stochastic Models of the lac Operon Genetic Network

The lac operon has been a paradigm for genetic regulation with positive feedback, and several modeling studies have described its dynamics at various levels of detail. However, it has not yet been analyzed how stochasticity can enrich the system's behavior, creating effects that are not observed in the deterministic case. To address this problem we use a comparative approach. We develop a reaction network for the dynamics of the lac operon genetic switch and derive corresponding deterministic and stochastic models that incorporate biological details. We then analyze the effects of key biomolecular mechanisms, such as promoter strength and binding affinities, on the behavior of the models. No assumptions or approximations are made when building the models other than those utilized in the reaction network. Thus, we are able to carry out a meaningful comparison between the predictions of the two models to demonstrate genuine effects of stochasticity. Such a comparison reveals that in the presence of stochasticity, certain biomolecular mechanisms can profoundly influence the region where the system exhibits bistability, a key characteristic of the lac operon dynamics. For these cases, the temporal asymptotic behavior of the deterministic model remains unchanged, indicating a role of stochasticity in modulating the behavior of the system.