Computational Systems Biology Research Articles

Efficient Emulation of Computer Models Utilising Multiple Known Boundaries of Differing Dimension

Emulation has been successfully applied across a wide variety of scientific disciplines for efficiently analysing computationally intensive models. We develop known boundary emulation strategies which utilise the fact that, for many computer models, there exist hyperplanes in the input parameter space for which the model output can be evaluated far more efficiently, whether this be analytically or just significantly faster using a more efficient and simpler numerical solver. The information contained on these known hyperplanes, or boundaries, can be incorporated into the emulation process via analytical update, thus involving no additional computational cost. In this article, we show that such analytical updates are available for multiple boundaries of various dimensions. We subsequently demonstrate which configurations of boundaries such analytical updates are available for, in particular by presenting a set of conditions that such a set of boundaries must satisfy. We demonstrate the powerful computational advantages of the known boundary emulation techniques developed on both an illustrative low-dimensional simulated example and a scientifically relevant and high-dimensional systems biology model of hormonal crosstalk in the roots of an Arabidopsis plant.

PhysiCOOL: A generalized framework for model Calibration and Optimization Of modeLing projects.

In silico models of biological systems are usually very complex and rely on a large number of parameters describing physical and biological properties that require validation. As such, parameter space exploration is an essential component of computational model development to fully characterize and validate simulation results. Experimental data may also be used to constrain parameter space (or enable model calibration) to enhance the biological relevance of model parameters. One widely used computational platform in the mathematical biology community is PhysiCell,which provides a standardized approach to agent-based models of biological phenomena at different time and spatial scales. Nonetheless, one limitation of PhysiCell is the lack of a generalized approach for parameter space exploration and calibration that can be run without high-performance computing access. Here, we present PhysiCOOL, an open-source Python library tailored to create standardized calibration and optimization routines for PhysiCell models.

Novel biomarkers for neoplastic progression from ulcerative colitis to colorectal cancer: a systems biology approach

In recent studies, the void of evaluation and in-depth understanding of unknown clinically relevant potential molecular biomarkers involved in colorectal cancer (CRC) from the inflammatory stage of ulcerative colitis (UC) to CRC metastasis, which can be suitable therapeutic targets, is deeply felt. The regulation and interaction among different cancer-promoting molecules, including messenger RNAs (mRNAs) and micro RNAs (miRNAs) in CRC and its progression, were the aim we pursued in this study. Using microarray data, we investigated the differential expression for five datasets, including mRNA and microRNA samples related to UC, tumor/normal. Then, using robust data analysis, separate lists of differentially expressed genes (DEGs) and differentially expressed miRNAs (DEmiRNAs) were identified, which were used for robust rank aggregation (RRA) and co-expression network analysis. Then, comprehensive computational systems biology analyses, including gene ontology and Kyoto encyclopedia of genes and genomic pathway enrichment analyses, mRNA-miRNA regulatory network, and survival analysis, were employed to achieve the aim of this study. Finally, we used clinical samples to validate this potential and new target. According to this systems biology approach, a total of 98 DEGs and 8 DEmiRNAs with common differential expression were identified. By combining the distinct results of RRA and network, several potential therapeutic targets, and predictive and prognostic biomarkers for UC and CRC were identified. These targets include six common hub genes, CXCL1, CXCL8, MMP7, SLCA16A9, PLAU, and TIMP1, which are upregulated. Among these, the important and new biomarker SLC16A9 is negatively regulated by hsa-mir-194-5p, and hsa-miR-378a-5p take. The findings of the present study provide new insight into the pathogenesis of CRC in UC. Our study suggests future evaluation of the functional role of SLC16A9 and hsa-mir-194-5p and hsa-miR-378a-5p in CRC development.

Cell-level systems biology model to study inflammatory bowel diseases and their treatment options.

To help understand the complex and therapeutically challenging inflammatory bowel diseases (IBDs), we developed a systems biology model of the intestinal immune system that is able to describe main aspects of IBD and different treatment modalities thereof. The model, including key cell types and processes of the mucosal immune response, compiles a large amount of isolated experimental findings from literature into a larger context and allows for simulations of different inflammation scenarios based on the underlying data and assumptions. In the context of a large and diverse virtual IBD population, we characterized the patients based on their phenotype (in contrast to healthy individuals, they developed persistent inflammation after a trigger event) rather than on a priori assumptions on parameter differences to a healthy individual. This allowed to reproduce the enormous diversity of predispositions known to lead to IBD. Analyzing different treatment effects, the model provides insight into characteristics of individual drug therapy. We illustrate for anti-TNF-α therapy, how the model can be used (i) to decide for alternative treatments with best prospects in the case of nonresponse, and (ii) to identify promising combination therapies with other available treatment options.

Whole-genome resequencing data support a single introduction of the invasive white pine sawfly, Diprion similis.

Biological introductions are unintended "natural experiments" that provide unique insights into evolutionary processes. Invasive phytophagous insects are of particular interest to evolutionary biologists studying adaptation, as introductions often require rapid adaptation to novel host plants. However, adaptive potential of invasive populations may be limited by reduced genetic diversity-a problem known as the "genetic paradox of invasions." One potential solution to this paradox is if there are multiple invasive waves that bolster genetic variation in invasive populations. Evaluating this hypothesis requires characterizing genetic variation and population structure in the invaded range. To this end, we assemble a reference genome and describe patterns of genetic variation in the introduced white pine sawfly, Diprion similis. This species was introduced to North America in 1914, where it has rapidly colonized the thin-needled eastern white pine (Pinus strobus), making it an ideal invasion system for studying adaptation to novel environments. To evaluate evidence of multiple introductions, we generated whole-genome resequencing data for 64 D. similis females sampled across the North American range. Both model-based and model-free clustering analyses supported a single population for North American D. similis. Within this population, we found evidence of isolation-by-distance and a pattern of declining heterozygosity with distance from the hypothesized introduction site. Together, these results support a single-introduction event. We consider implications of these findings for the genetic paradox of invasion and discuss priorities for future research in D. similis, a promising model system for invasion biology.

Investigating the comorbidity of COPD and tuberculosis, a computational study

Recent research has shown that people who suffer from chronic obstructive pulmonary disease (COPD) have a greater propensity to contract and develop tuberculosis (TB) than the general population. Not only is the hazard ratio for contracting active tuberculosis triple that of the general population for those with COPD, but that the probability of death from any cause during the first year was double that of the tuberculosis population as a whole. This observation suggests that patients with COPD are less likely to progress to latent tuberculosis infection (LTBI) and are more likely to develop active tuberculosis than the general population. While similar susceptibility rates to TB are known to occur in populations with other ailments of the lung, particularly HIV, emphysema or asthma, patients with COPD (both emphysema and chronic bronchitis) are statistically more at risk for the disease. To examine the comorbidity effects of COPD on tuberculosis disease and granuloma formation, the process by which Mycobacterium tuberculosis (Mtb) is either contained or disseminates, we used a multi-scale model that integrates pathophysiological and immunopathological aspects of COPD and TB. Depicting chronic obstructive pulmonary disease smoker and non-smoker populations, we integrate agent-based models (ABM) of cellular immune response, physiological models of pulmonary capacity for COPD smoker/non-smoker, systems biology models of macrophage immune response to Mtb, and metabolic models to capture intracellular and extracellular Mtb metabolism and proliferation. We use our model to investigate key drivers of disease outcomes of clearance, granuloma-based containment, and disseminated disease in individuals with COPD and TB for smoking and non-smoking populations.

Cellular nitric oxide synthesis is affected by disorders in the interdependent [Formula: see text] and [Formula: see text] dynamics during cystic fibrosis disease.

Calcium ([Formula: see text]), inositol trisphosphate ([Formula: see text]), and nitric oxide (NO) signaling are essential to maintain the structural integrity and physiological activity of fibroblast cells. The accumulation of excess quantity of NO for longer periods can lead to a variety of fibrotic disorders, including heart disease, penile fibrosis in Peyronie's disease (PD), and cystic fibrosis. The dynamics of these three signaling processes and their interdependence in fibroblast cells are not clearly known to date. A systems biology model is proposed using reaction-diffusion equations for calcium, [Formula: see text], and calcium-dependent NO synthesis in fibroblast cells. The finite element method (FEM) is used to examine [Formula: see text], [Formula: see text], and NO regulation and dysregulation in cells. The results throw light on the conditions that disturb the coupled [Formula: see text] and [Formula: see text] dynamics and the influence of these factors on the levels of NO concentration in the fibroblast cell. The findings suggest that changes in source inflow, buffers, and diffusion coefficient might induce an increase or reduction in nitric oxide and [Formula: see text] synthesis, resulting in fibroblast cell diseases. Furthermore, the findings provide new information regarding the size and intensity of diseases in response to changes in several factors of their dynamics, which has been linked to the development of cystic fibrosis and cancer. This knowledge could be valuable for developing novel approaches to the diagnosis of diseases and therapies for various disorders of fibroblast cells.

Proteomic quantification of receptor tyrosine kinases involved in the development and progression of colorectal cancer liver metastasis.

Alterations in expression and activity of human receptor tyrosine kinases (RTKs) are associated with cancer progression and in response to therapeutic intervention. Thus, protein abundance of 21 RTKs was assessed in 15 healthy and 18 cancerous liver samples [2 primary and 16 colorectal cancer liver metastasis (CRLM)] matched with non-tumorous (histologically normal) tissue, by a validated QconCAT-based targeted proteomic approach. It was demonstrated, for the first time, that the abundance of EGFR, INSR, VGFR3 and AXL, is lower in tumours relative to livers from healthy individuals whilst the opposite is true for IGF1R. EPHA2 was upregulated in tumour compared with histologically normal tissue surrounding it. PGFRB levels were higher in tumours relative to both histologically normal tissue surrounding tumour and tissues taken from healthy individuals. The abundances of VGFR1/2, PGFRA, KIT, CSF1R, FLT3, FGFR1/3, ERBB2, NTRK2, TIE2, RET, and MET were, however, comparable in all samples. Statistically significant, but moderate correlations were observed (Rs > 0.50, p < 0.05) for EGFR with INSR and KIT. FGFR2 correlated with PGFRA and VGFR1 with NTRK2 in healthy livers. In non-tumorous (histologically normal) tissues from cancer patients, there were correlations between TIE2 and FGFR1, EPHA2 and VGFR3, FGFR3 and PGFRA (p < 0.05). EGFR correlated with INSR, ERBB2, KIT and EGFR, and KIT with AXL and FGFR2. In tumours, CSF1R correlated with AXL, EPHA2 with PGFRA, and NTRK2 with PGFRB and AXL. Sex, liver lobe and body mass index of donors had no impact on the abundance of RTKs, although donor age showed some correlations. RET was the most abundant of these kinases in non-tumorous tissues (~35%), while PGFRB was the most abundant RTK in tumours (~47%). Several correlations were also observed between the abundance of RTKs and proteins relevant to drug pharmacokinetics (enzymes and transporters). DiscussionThis study quantified perturbation to the abundance of several RTKs in cancer and the value generated in this study can be used as input to systems biology models defining liver cancer metastases and biomarkers of its progression.

Bayesian estimation reveals that reproducible models in Systems Biology get more citations

The Systems Biology community has taken numerous actions to develop data and modeling standards towards FAIR data and model handling. Nevertheless, the debate about incentives and rewards for individual researchers to make their results reproducible is ongoing. Here, we pose the specific question of whether reproducible models have a higher impact in terms of citations. Therefore, we statistically analyze 328 published models recently classified by Tiwari et al. based on their reproducibility. For hypothesis testing, we use a flexible Bayesian approach that provides complete distributional information for all quantities of interest and can handle outliers. The results show that in the period from 2013, i.e., 10 years after the introduction of SBML, to 2020, the group of reproducible models is significantly more cited than the non-reproducible group. We show that differences in journal impact factors do not explain this effect and that this effect increases with additional standardization of data and error model integration via PEtab. Overall, our statistical analysis demonstrates the long-term merits of reproducible modeling for the individual researcher in terms of citations. Moreover, it provides evidence for the increased use of reproducible models in the scientific community.

High-speed imaging of giant unilamellar vesicle production in cDICE.

Giant unilamellar vesicles (GUVs) are cell-sized containers that are commonly used as three-dimensional model membranes in biophysics, as in vitro model systems in synthetic biology, and even as cargo carriers in various other research fields. Despite their ubiquitous use, there is still no one-size-fits-all GUV production method. Over the years, numerous methods have been developed, attempting to meet the demanding requirements of robustness, reliability, and high yield while simultaneously achieving robust encapsulation. Double emulsion-based methods are often praised for their apparent simplicity and good yields; hence, methods like continuous droplet interface crossing encapsulation (cDICE) that make use of this principle, have gained popularity in recent years. In cDICE, aqueous droplets that originate from a capillary orifice are continuously forced through an oil-water interface by centrifugal force, thereby forming a lipid bilayer and thus GUVs. Although cDICE and related methods are frequently used in the field, the complexity of the underlying principles and fluid dynamics has not been considered previously, and how exactly the GUVs are being formed remains unknown. To elucidate the process of GUV formation in cDICE, we have developed a high-speed microscopy setup that allows us to visualize GUV formation in real time. We focused on the capillary orifice, where initial droplet formation occurs, and on the oil-water interface, where droplets are converted into GUVs. Our experiments reveal a complex droplet formation process at the capillary orifice and suboptimal droplet transfer through the water-oil interface, which we explain using fluid dynamics and theoretical modeling. Our results are a first step towards explaining the widely observed variation in encapsulation efficiency and size polydispersity in cDICE. Ultimately, these results will contribute to a better understanding of GUV formation processes in cDICE and in extension, in double emulsion-based methods in general.

Rule-based modelling of biological systems using regulated rewriting

In systems biology, models play a crucial role in understanding studied systems. There are many modelling approaches, among which rewriting systems provide a framework for describing systems on a mechanistic level. Describing biochemical processes often requires incorporating knowledge on an abstract level to simplify the system description or substitute the missing details. For this purpose, we present regulation mechanisms, an extension of this formalism with additional controls on the rewriting process. We introduce several regulation mechanisms and apply them to a rule-based language, a notation suitable for modelling biological phenomena. Finally, we demonstrate the usage of such regulations on several case studies from the biochemical domain.

Autocorrelated measurement processes and inference for ordinary differential equation models of biological systems

Ordinary differential equation models are used to describe dynamic processes across biology. To perform likelihood-based parameter inference on these models, it is necessary to specify a statistical process representing the contribution of factors not explicitly included in the mathematical model. For this, independent Gaussian noise is commonly chosen, with its use so widespread that researchers typically provide no explicit justification for this choice. This noise model assumes ‘random’ latent factors affect the system in the ephemeral fashion resulting in unsystematic deviation of observables from their modelled counterparts. However, like the deterministically modelled parts of a system, these latent factors can have persistent effects on observables. Here, we use experimental data from dynamical systems drawn from cardiac physiology and electrochemistry to demonstrate that highly persistent differences between observations and modelled quantities can occur. Considering the case when persistent noise arises owing only to measurement imperfections, we use the Fisher information matrix to quantify how uncertainty in parameter estimates is artificially reduced when erroneously assuming independent noise. We present a workflow to diagnose persistent noise from model fits and describe how to remodel accounting for correlated errors.

Mechanistic Understanding of D-Glucaric Acid to Support Liver Detoxification Essential to Muscle Health Using a Computational Systems Biology Approach.

Liver and muscle health are intimately connected. Nutritional strategies that support liver detoxification are beneficial to muscle recovery. Computational-in silico-molecular systems' biology analysis of supplementation of calcium and potassium glucarate salts and their metabolite D-glucaric acid (GA) reveals their positive effect on mitigation of liver detoxification via four specific molecular pathways: (1) ROS production, (2) deconjugation, (3) apoptosis of hepatocytes, and (4) β-glucuronidase synthesis. GA improves liver detoxification by downregulating hepatocyte apoptosis, reducing glucuronide deconjugates levels, reducing ROS production, and inhibiting β-Glucuronidase enzyme that reduces re-absorption of toxins in hepatocytes. Results from this in silico study provide an integrative molecular mechanistic systems explanation for the mitigation of liver toxicity by GA.

Characterization of Integrase and Excisionase Activity in a Cell-Free Protein Expression System Using a Modeling and Analysis Pipeline.

We present a full-stack modeling, analysis, and parameter identification pipeline to guide the modeling and design of biological systems starting from specifications to circuit implementations and parametrizations. We demonstrate this pipeline by characterizing the integrase and excisionase activity in a cell-free protein expression system. We build on existing Python tools─BioCRNpyler, AutoReduce, and Bioscrape─to create this pipeline. For enzyme-mediated DNA recombination in a cell-free system, we create detailed chemical reaction network models from simple high-level descriptions of the biological circuits and their context using BioCRNpyler. We use Bioscrape to show that the output of the detailed model is sensitive to many parameters. However, parameter identification is infeasible for this high-dimensional model; hence, we use AutoReduce to automatically obtain reduced models that have fewer parameters. This results in a hierarchy of reduced models under different assumptions to finally arrive at a minimal ODE model for each circuit. Then, we run sensitivity analysis-guided Bayesian inference using Bioscrape for each circuit to identify the model parameters. This process allows us to quantify integrase and excisionase activity in cell extracts enabling complex-circuit designs that depend on accurate control over protein expression levels through DNA recombination. The automated pipeline presented in this paper opens up a new approach to complex circuit design, modeling, reduction, and parametrization.

Integrated regulatory and metabolic networks of the tumor microenvironment for therapeutic target prioritization.

Translation of genomic discovery, such as single-cell sequencing data, to clinical decisions remains a longstanding bottleneck in the field. Meanwhile, computational systems biological models, such as cellular metabolism models and cell signaling pathways, have emerged as powerful approaches to provide efficient predictions in metabolites and gene expression levels, respectively. However, there has been limited research on the integration between these two models. This work develops a methodology for integrating computational models of probabilistic gene regulatory networks with a constraint-based metabolism model. By using probabilistic reasoning with Bayesian Networks, we aim to predict cell-specific changes under different interventions, which are embedded into the constraint-based models of metabolism. Applications to single-cell sequencing data of glioblastoma brain tumors generate predictions about the effects of pharmaceutical interventions on the regulatory network and downstream metabolisms in different cell types from the tumor microenvironment. The model presents possible insights into treatments that could potentially suppress anaerobic metabolism in malignant cells with minimal impact on other cell types' metabolism. The proposed integrated model can guide therapeutic target prioritization, the formulation of combination therapies, and future drug discovery. This model integration framework is also generalizable to other applications, such as different cell types, organisms, and diseases.

Multi-omics data integration methods and their applications in psychiatric disorders

To study mental illness and health, in the past researchers have often broken down their complexity into individual subsystems (e.g., genomics, transcriptomics, proteomics, clinical data) and explored the components independently. Technological advancements and decreasing costs of high throughput sequencing has led to an unprecedented increase in data generation. Furthermore, over the years it has become increasingly clear that these subsystems do not act in isolation but instead interact with each other to drive mental illness and health. Consequently, individual subsystems are now analysed jointly to promote a holistic understanding of the underlying biological complexity of health and disease. Complementing the increasing data availability, current research is geared towards developing novel methods that can efficiently combine the information rich multi-omics data to discover biologically meaningful biomarkers for diagnosis, treatment, and prognosis. However, clinical translation of the research is still challenging. In this review, we summarise conventional and state-of-the-art statistical and machine learning approaches for discovery of biomarker, diagnosis, as well as outcome and treatment response prediction through integrating multi-omics and clinical data. In addition, we describe the role of biological model systems and in silico multi-omics model designs in clinical translation of psychiatric research from bench to bedside. Finally, we discuss the current challenges and explore the application of multi-omics integration in future psychiatric research. The review provides a structured overview and latest updates in the field of multi-omics in psychiatry.

The Influence of 3-D Model Biological Systems in Understanding Cellular Diversity

This study focused on the influence of instructional 3D Model Biological Systems (MBS) in understanding cell diversity. It determined the academic performance of the students according to content knowledge, quality of the artwork, and communication skills; and determined what group of learners benefited most from using 3D instructional models. It involved 30 second-year high school students in a private school in Imus, Cavite. Data were treated statistically with arithmetic mean, dependent t-test, gain scores, Pearson r Product Moment of Correlation, one-way analysis of variance (ANOVA), F-test, and Tukey multiple comparison test. Findings revealed that the use of 3D instructional cell models was effective in improving the academic performance of the students in terms of content knowledge, quality of the artwork, and communication skills. The computed mean difference of 3.60 and t-value of 10.998 for 29 degrees of freedom, which is 2.756 at a .01 level of confidence, showed a significant difference from their pre-test and post-test scores. This implies that participants obtained much learning on cellular diversity using 3D cell models. All groups of learners benefited most from using 3D instructional cell models thus students’ academic performance in Biology has improved. Thus, the study recommends that 3D instructional models should be used in teaching professional Science subjects such as Biology. More so, similar studies should also be conducted not only in Biology subject but also in Chemistry and Physics courses. Other 3D Model Biological Systems can be used for instruction to Senior High School Science majors and college students. Keywords: Cellular Diversity, 3D Model Biological Systems, General Biology, Cell Models

Hyperbolic matrix factorization improves prediction of drug-target associations

Past research in computational systems biology has focused more on the development and applications of advanced statistical and numerical optimization techniques and much less on understanding the geometry of the biological space. By representing biological entities as points in a low dimensional Euclidean space, state-of-the-art methods for drug-target interaction (DTI) prediction implicitly assume the flat geometry of the biological space. In contrast, recent theoretical studies suggest that biological systems exhibit tree-like topology with a high degree of clustering. As a consequence, embedding a biological system in a flat space leads to distortion of distances between biological objects. Here, we present a novel matrix factorization methodology for drug-target interaction prediction that uses hyperbolic space as the latent biological space. When benchmarked against classical, Euclidean methods, hyperbolic matrix factorization exhibits superior accuracy while lowering embedding dimension by an order of magnitude. We see this as additional evidence that the hyperbolic geometry underpins large biological networks.

Learning the structure of the mTOR protein signaling pathway from protein phosphorylation data

Statistical learning of the structures of cellular networks, such as protein signaling pathways, is a topical research field in computational systems biology. To get the most information out of experimental data, it is often required to develop a tailored statistical approach rather than applying one of the off-the-shelf network reconstruction methods. The focus of this paper is on learning the structure of the mTOR protein signaling pathway from immunoblotting protein phosphorylation data. Under two experimental conditions eleven phosphorylation sites of eight key proteins of the mTOR pathway were measured at ten non-equidistant time points. For the statistical analysis we propose a new advanced hierarchically coupled non-homogeneous dynamic Bayesian network (NH-DBN) model, and we consider various data imputation methods for dealing with non-equidistant temporal observations. Because of the absence of a true gold standard network, we propose to use predictive probabilities in combination with a leave-one-out cross validation strategy to objectively cross-compare the accuracies of different NH-DBN models and data imputation methods. Finally, we employ the best combination of model and data imputation method for predicting the structure of the mTOR protein signaling pathway.

A Mori–Zwanzig Dissipative Particle Dynamics Approach for Anisotropic Coarse Grained Molecular Dynamics

Coarse grained (CG) molecular dynamics simulations are widely used to accelerate atomistic simulations but generally lack a formalism to preserve the dynamics of the system. For spherical particles, the Mori-Zwanzig approach, while computationally complex, has ameliorated this problem. Here we present an anisotropic dissipative particle dynamics (ADPD) model as an extension of this approach, which accounts for the anisotropy for both conservative and nonconservative interactions. For a simple anisotropic system we parametrize the coarse grained force field representing ellipsoidal CG particles from the full-atomistic simulation. To represent the anisotropy of the system, both the conservative and dissipative terms are approximated using the Gay-Berne (GB) functional forms via a force-matching approach. We compare our model with other CG models and demonstrate that it yields better results in both static and dynamical properties. The inclusion of the anisotropic nonconservative force preserves the microscopic dynamical details, and hence the dynamical properties, such as diffusivity, can be better reproduced by the aspherical model. By generalizing the isotropic DPD model, this framework is effective and promising for the development of the CG model for polymers, macromolecules, and biological systems.