Medical datasets are typically affected by issues such as missing values, class imbalance, a heterogeneous feature types, and a high number of features versus a relatively small number of samples, preventing machine learning models from obtaining proper results in classification and regression tasks. This paper introduces AutoML-Med, an Automated Machine Learning tool specifically designed to address these challenges, minimizing user intervention and identifying the optimal combination of preprocessing techniques and predictive models. AutoML-Med's architecture incorporates Latin Hypercube Sampling (LHS) for exploring preprocessing methods, trains models using selected metrics, and utilizes Partial Rank Correlation Coefficient (PRCC) for fine-tuned optimization of the most influential preprocessing steps. Experimental results demonstrate AutoML-Med's effectiveness in two different clinical settings, achieving higher balanced accuracy and sensitivity, which are crucial for identifying at-risk patients, compared to other state-of-the-art tools. AutoML-Med's ability to improve prediction results, especially in medical datasets with sparse data and class imbalance, highlights its potential to streamline Machine Learning applications in healthcare.

Agent-Based Modeling and Simulation (ABMS) has been increasingly applied in various research fields, thanks to the capability of these models to describe fine-grained realworld behavior and to the ease of interpretation by domain experts. However, such models lack a formal definition and well-defined semantics that are common to the different tools supporting ABMS. This may occasionally lead to greater complexity in interpreting the results with respect to other modeling approaches. To address this issue, an ABM semantics that adopts a continuous-time approach and a next-event time advance simulation algorithm has been formally defined and presented. Such an approach may lead to high computation times as it requires recalculations of activity rates for all the agents after each event. In this preliminary study, we exploit the FLAME GPU framework to evaluate the benefits that GPU computing may bring to the performance of our simulation algorithm.

We present here the results obtained by applying the Universal Immune System Simulator (UISS) computational platform in one declinations developed for COVID19 disease (UISS-COVID19) to model the efficacy, demonstrated during the SAVE-MORE study, of anakinra drug on patients affected by a severe illness induced by SARS-COV-2. The goal is to further validate the model capabilities and to confirm through computer simulations the biological hypotheses on the mechanisms of actions involved with the use of anakinra drug. UISS, and in silico models in general, have the potential to be integrated in future clinical trial pipelines and to be adopted as technologies to support clinical trials to evaluate the effectiveness of any drug or or xenobiotic.

Open science and reproducibility are two key pillars of modern scientific research. Open science is making scientific research and data accessible and transparent to the broader scientific community and the public. Reproducibility, on the other hand, is the ability to replicate and confirm research results by following the same methods and procedures. Reproducibility is thus crucial because it ensures the reliability and validity of scientific findings. The relationship between open science and reproducibility is intertwined; indeed open science practices, such as sharing raw data, detailed methodologies, and code, greatly facilitate the reproducibility of research. In recent years, concerns about the reproducibility of scientific research have gained prominence, and indeed scientists still lament the lack of details in the methods sections of published papers and the unavailability of raw data from the authors.To assist cellular biologists and immunologists and to promote a more transparent, open and reproducible research practice, we developed OmniReproducibleCellAnalysis (ORCA), a new Shiny Application based in R, for the semi-automated analysis of Western Blot (WB), Reverse Transcription-quantitative PCR (RT-qPCR), Enzyme-Linked ImmunoSorbent Assay (ELISA), Endocytosis and Cytotoxicity experiments. ORCA is open-source and approachable by scientists without advanced R language knowledge. Our application automatically compiles a report containing the finalized data analysis and all its preliminary and intermediate steps, ensuring data analysis standardization and reproducibility. Furthermore, ORCA allows to upload raw data and results directly on the data repository Harvard Dataverse, a valuable tool for promoting transparency and data accessibility in scientific research.By employing ORCA, scientists will cut down analysis time and human-dependent errors, while taking a step towards a research practice compliant with Open Science and FAIR principle.

The Universal Immune System Simulator (UISS) is a computational framework based on agent-based modelling (ABM) paradigm that has been specifically developed for simulating the immune system behaviour in presence of diseases and treatments. It has a long history of development, ranging from its initial applications into the field of tumor immunology and then moving towards wide disease modelling scenarios such as influenza, Multiple Sclerosis and atherosclerosis. Recently, inside the STriTuVaD H2020 EU project, it has been specialized to simulate tuberculosis dynamics and its interaction with the immune system, including the efficacy of the combined action of various treatments such as isoniazid and novel vaccines. TB simulation entitles large scale (e.g., tissue to organ scale) simulations over a wide digital population cohort. The computational costs of running large scale simulations are prohibitive using traditional forms of CPU simulation. This paper considers the use of parallel to gpu-based computing approaches via an agent-based domain independent complex systems simulator, FLAME GPU. Integration of FLAME GPU with UISS enables the simulation of larger, more complex problem domains. The combined UISS-FLAMEGPU simulator provides vastly increased performance characteristics for large problems, with a speedup of 4.22x for a typical tuberculosis model simulating 128 microlitres. FLAME GPU abstracts away a significant portion of the normal programming that would be required to effectively parallelise a model of this complexity. Adaptations were made to increase performance, such as message mutation and parallelisation of certain algorithms.

The immune system (IS) represents a complex network of cells and molecules devoted to the protection of individuals from external pathogens, and in terms of complexity, it is only second to the central nervous system. As our knowledge of the IS mechanisms has become more exhaustive, interest has grown in applying modeling and simulation techniques in this context. In particular, among these techniques, the Agent Based Models (ABMs) have been increasingly applied for the IS simulation. One of the major drawbacks of ABMs is represented by the lack of well-defined semantics, which may lead to inconsistent results in comparison to other stochastic approaches. In this paper, we make use of the well-defined semantics and the simulation algorithm for ABMs that we proposed in [1] to implement a few models of the Cancer-Immune System. Comparing ABMs and Gillespie’s Stochastic Simulation Algorithm results we show that our methodology brings coherence among the results of ABMs and SSA.

BACKGROUND: Nowadays, the inception of computer modeling and simulation in life science is a matter of fact. This is one of the reasons why regulatory authorities are open in considering in silico trials evidence for the assessment of safeness and efficacy of medicinal products. In this context, mechanistic Agent-Based Models are increasingly used. Unfortunately, there is still a lack of consensus in the verification assessment of Agent-Based Models for regulatory approval needs. VV&UQ is an ASME standard specifically suited for the verification, validation, and uncertainty quantification of medical devices. However, it can also be adapted for the verification assessment of in silico trials for medicinal products. RESULTS: Here, we propose a set of automatic tools for the mechanistic Agent-Based Model verification assessment. As a working example, we applied the verification framework to an Agent-Based Model in silico trial used in the COVID-19 context. CONCLUSIONS: Using the described verification computational workflow allows researchers and practitioners to easily perform verification steps to prove Agent-Based Models robustness and correctness that provide strong evidence for further regulatory requirements.

The application of Agent Based Models (ABMs) in biology and immunology has recently come to the fore, thanks to their ability to accurately describe complex biological behaviors, rules, and interactions, without the need to use complex mathematical formalisms. However, even if there is a growing interest in applying such methodologies to improve and speed up the research of novel pharmaceutical products, verification and validation procedures voted at assessing ABMs credibility are far from being well-established. We present Verify, the first toolbox of instruments selected and designed for the verification of discrete-time models, with a focus on agent-based approaches. The toolbox has a friendly GUI, does not require the installation of any additional software, and can easily find possible numerical errors and incongruences that may affect such models.

An ever-increasing number of tools and databases offers pieces of evidence and knowledge about gene function annotations, protein interactions, and experimentally validated biological pathways. These resources represent an excellent and essential instrument to facilitate pathway analysis. Among them, particular attention should be given to the one capable of finding alternative pathways or cross-talks events and potential drug target candidates. Here, we evaluated PETAL predictive capability, a Python tool that automatically explores and scans the relevant nodes within a KEGG pathway. Starting from three specific cancer scenarios (chronic myelogenous leukemia, non-small cell lung cancer, and head and neck squamous cell carcinoma) and related literature results about potential driver genes of drug resistance to EGFR tyrosine kinases inhibitors (SNCA, BCL-6 and YAP-1), we used PETAL to test its capability to identify in parallel these potential target genes involved in tumor progression and EGFR inhibitors resistance. By searching in-depth for ancestor and descendent nodes of SNCA, BCL-6 and YAP-1, across the EGFR pathway, we found that PETAL was able to detect the same targets investigated in the recent literature. Finally, this retrospective work emphasizes that PETAL could represent a powerful tool that can be used to improve the understanding of complex biological pathways and speed-up the identification of potential biomarkers and therapeutical candidates in cancer potentially in any other disease.

BACKGROUND: The STriTuVaD project, funded by Horizon 2020, aims to test through a Phase IIb clinical trial one of the most advanced therapeutic vaccines against tuberculosis. As part of this initiative, we have developed a strategy for generating in silico patients consistent with target population characteristics, which can then be used in combination with in vivo data on an augmented clinical trial. RESULTS: One of the most challenging tasks for using virtual patients is developing a methodology to reproduce biological diversity of the target population, ie, providing an appropriate strategy for generating libraries of digital patients. This has been achieved through the creation of the initial immune system repertoire in a stochastic way, and through the identification of a vector of features that combines both biological and pathophysiological parameters that personalise the digital patient to reproduce the physiology and the pathophysiology of the subject. CONCLUSIONS: We propose a sequential approach to sampling from the joint features population distribution in order to create a cohort of virtual patients with some specific characteristics, resembling the recruitment process for the target clinical trial, which then can be used for augmenting the information from the physical the trial to help reduce its size and duration.

BACKGROUND: Multiple Sclerosis (MS) represents nowadays in Europe the leading cause of non-traumatic disabilities in young adults, with more than 700,000 EU cases. Although huge strides have been made over the years, MS etiology remains partially unknown. Furthermore, the presence of various endogenous and exogenous factors can greatly influence the immune response of different individuals, making it difficult to study and understand the disease. This becomes more evident in a personalized-fashion when medical doctors have to choose the best therapy for patient well-being. In this optics, the use of stochastic models, capable of taking into consideration all the fluctuations due to unknown factors and individual variability, is highly advisable. RESULTS: We propose a new model to study the immune response in relapsing remitting MS (RRMS), the most common form of MS that is characterized by alternate episodes of symptom exacerbation (relapses) with periods of disease stability (remission). In this new model, both the peripheral lymph node/blood vessel and the central nervous system are explicitly represented. The model was created and analysed using Epimod, our recently developed general framework for modeling complex biological systems. Then the effectiveness of our model was shown by modeling the complex immunological mechanisms characterizing RRMS during its course and under the DAC administration. CONCLUSIONS: Simulation results have proven the ability of the model to reproduce in silico the immune T cell balance characterizing RRMS course and the DAC effects. Furthermore, they confirmed the importance of a timely intervention on the disease course.

BACKGROUND: SARS-CoV-2 is a severe respiratory infection that infects humans. Its outburst entitled it as a pandemic emergence. To get a grip on this outbreak, specific preventive and therapeutic interventions are urgently needed. It must be said that, until now, there are no existing vaccines for coronaviruses. To promptly and rapidly respond to pandemic events, the application of in silico trials can be used for designing and testing medicines against SARS-CoV-2 and speed-up the vaccine discovery pipeline, predicting any therapeutic failure and minimizing undesired effects. RESULTS: We present an in silico platform that showed to be in very good agreement with the latest literature in predicting SARS-CoV-2 dynamics and related immune system host response. Moreover, it has been used to predict the outcome of one of the latest suggested approach to design an effective vaccine, based on monoclonal antibody. Universal Immune System Simulator (UISS) in silico platform is potentially ready to be used as an in silico trial platform to predict the outcome of vaccination strategy against SARS-CoV-2. CONCLUSIONS: In silico trials are showing to be powerful weapons in predicting immune responses of potential candidate vaccines. Here, UISS has been extended to be used as an in silico trial platform to speed-up and drive the discovery pipeline of vaccine against SARS-CoV-2.

BACKGROUND: In 2018, about 10 million people were found infected by tuberculosis, with approximately 1.2 million deaths worldwide. Despite these numbers have been relatively stable in recent years, tuberculosis is still considered one of the top 10 deadliest diseases worldwide. Over the years, Mycobacterium tuberculosis has developed a form of resistance to first-line tuberculosis treatments, specifically to isoniazid, leading to multi-drug-resistant tuberculosis. In this context, the EU and Indian DBT funded project STriTuVaD-In Silico Trial for Tuberculosis Vaccine Development-is supporting the identification of new interventional strategies against tuberculosis thanks to the use of Universal Immune System Simulator (UISS), a computational framework capable of predicting the immunity induced by specific drugs such as therapeutic vaccines and antibiotics. RESULTS: Here, we present how UISS accurately simulates tuberculosis dynamics and its interaction within the immune system, and how it predicts the efficacy of the combined action of isoniazid and RUTI vaccine in a specific digital population cohort. Specifically, we simulated two groups of 100 digital patients. The first group was treated with isoniazid only, while the second one was treated with the combination of RUTI vaccine and isoniazid, according to the dosage strategy described in the clinical trial design. UISS-TB shows to be in good agreement with clinical trial results suggesting that RUTI vaccine may favor a partial recover of infected lung tissue. CONCLUSIONS: In silico trials innovations represent a powerful pipeline for the prediction of the effects of specific therapeutic strategies and related clinical outcomes. Here, we present a further step in UISS framework implementation. Specifically, we found that the simulated mechanism of action of RUTI and INH are in good alignment with the results coming from past clinical phase IIa trials.

Different research communities have developed various approaches to assess the credibility of predictive models. Each approach usually works well for a specific type of model, and under some epistemic conditions that are normally satisfied within that specific research domain. Some regulatory agencies recently started to consider evidences of safety and efficacy on new medical products obtained using computer modelling and simulation (which is referred to as In Silico Trials); this has raised the attention in the computational medicine research community on the regulatory science aspects of this emerging discipline. But this poses a foundational problem: in the domain of biomedical research the use of computer modelling is relatively recent, without a widely accepted epistemic framing for model credibility. Also, because of the inherent complexity of living organisms, biomedical modellers tend to use a variety of modelling methods, sometimes mixing them in the solution of a single problem. In such context merely adopting credibility approaches developed within other research communities might not be appropriate. In this paper we propose a theoretical framing for assessing the credibility of a predictive models for In Silico Trials, which accounts for the epistemic specificity of this research field and is general enough to be used for different type of models.

Intracellular biochemical reactions emerge from the interaction among multiple extracellular signaling components. Considering the number, type and connections of the signaling components represents a needed step to characterize, identify and describe potential targets for a clinical purpose. However, it is increasingly documented that the presence of sub-types of signaling proteins, branching and crosstalk may lead to very variable outcomes in the same path, which is not always well defined experimentally. For this reason, we present an improved version of the algorithm based on the MapReduce paradigm to facilitate the discovery of new therapeutic targets. Our algorithm allows you to scan and perform a search in depth of biological pathway in order to analyze less recurrent and therefore non-trivial paths. These routes represent a chain of biochemical interactions among different biological actors that can be represented by quite distant nodes along the pathway. This type of analysis can also be performed manually, but with high execution times due to the large amount of pathways and genes present. Thus, our tool performs exhaustive analysis in an automated way, drastically reducing the time required. Our proposal allows us to discover the genes far from the initial target genes, also showing the number of occurrences of a given path found within the set of biological pathways analyzed during the simulation.

EC funded STriTuVaD project aims to test, through a phase IIb clinical trial, two of the most advanced therapeutic vaccines against tuberculosis. In parallel, we have extended the Universal Immune System Simulator to include all relevant determinants of such clinical trial, to establish its predictive accuracy against the individual patients recruited in the trial, to use it to generate digital patients and predict their response to the HRT being tested, and to combine them to the observations made on physical patients using a new in silico-augmented clinical trial approach that uses a Bayesian adaptive design. This approach, where found effective could drastically reduce the cost of innovation in this critical sector of public healthcare. One of the most challenging task is to develop a methodology to reproduce biological diversity of the subjects that have to be simulated, i.e., provide an appropriate strategy for the generation of libraries of digital patients. This has been achieved through the the creation of the initial immune system repertoire in a stochastic way, and though the identification of a “vector of features” that combines both biological and pathophysiological parameters that personalize the digital patient to reproduce the physiology and the pathophysiology of the subject.

It is well known that the response of individuals to disease varies, either because of unpredictable exogenous events, such as possibly unknown environmental effects, or just because of endogenous factors, i.e. different genetic background. In particular, when a treatment effectiveness has to be validated, the individual variability should be taken into account by exploiting stochastic models. Relapsing Remitting Multiple Sclerosis (RRMS) is an unpredictable and complex disease, whose random behaviour perfectly fits the study with stochastic models. RRMS is the most common form of Multiple Sclerosis (MS), an immune-mediated inflammatory disease of the central nervous system, characterized by alternate episodes of symptom exacerbation (relapses) with periods of disease stability (remission). Several treatments were proposed to contrast the disease progression. Among these, Daclizumab initially exhibited promising results. However, due to the risk of serious side effects the treatment has been retired. We propose a stochastic and an hybrid extension, based on a generalization of the high level Petri Net formalism, of an existing model of Daclizumab effects on RRMS. The model is developed to investigate the complex mechanisms and unpredictable behaviour characterizing the RRMS disease and its relapsing, especially under the Daclizumab administration.

Tuberculosis (TB) is one of the deadliest diseases worldwide, with 1,5 million fatalities every year along with potential devastating effects on society, families and individuals. To address this alarming burden, vaccines can play a fundamental role, even though to date no fully effective TB vaccine really exists. Current treatments involve several combinations of antibiotics administered to TB patients for up to two years, leading often to financial issues and reduced therapy adherence. Along with this, the development and spread of drug-resistant TB strains is another big complicating matter. Faced with these challenges, there is an urgent need to explore new vaccination strategies in order to boost immunity against tuberculosis and shorten the duration of treatment. Computational modeling represents an extraordinary way to simulate and predict the outcome of vaccination strategies, speeding up the arduous process of vaccine pipeline development and relative time to market. Here, we present EU - funded STriTuVaD project computational platform able to predict the artificial immunity induced by RUTI and ID93/GLA-SE, two specific tuberculosis vaccines. Such an in silico trial will be validated through a phase 2b clinical trial. Moreover, STriTuVaD computational framework is able to inform of the reasons for failure should the vaccinations strategies against M. tuberculosis under testing found not efficient, which will suggest possible improvements.

Thyroid cancer is the commonest endocrine malignancy. Mutation in the BRAF serine/threonine kinase is the most frequent genetic alteration in thyroid cancer. Target therapy for advanced and poorly differentiated thyroid carcinomas include BRAF pathway inhibitors. Here, we evaluated the role of MAP3K8 expression as a potential driver of resistance to BRAF inhibition in thyroid cancer. By analyzing Gene Expression Omnibus data repository, across all thyroid cancer histotypes, we found that MAP3K8 is up-regulated in poorly differentiated thyroid carcinomas and its expression is related to a stem cell like phenotype and a poorer prognosis and survival. Taken together these data unravel a novel mechanism for thyroid cancer progression and chemo-resistance and confirm previous results obtained in cultured thyroid cancer stem cells.

We present here a novel algorithm based on a MapReduce approach to facilitate the discovery of novel therapeutic targets. The proposed algorithm has been enabled to scan a set biological pathways in order to discover non-trivial (less common) routes. Such routes represent a chain of biochemical interactions among different biological actors. These actors can be represented by quite distant nodes along the devised pathway. Our approach detects nodes that are far from the initial target nodes, also showing the number of times that a given route has been found inside the selected set of biological pathways.

MOTIVATION: Val600Glu (V600E) mutation is the most common BRAF mutation detected in thyroid cancer. Hence, recent research efforts have been performed trying to explore several inhibitors of the V600E mutation-containing BRAF kinase as potential therapeutic options in thyroid cancer refractory to standard interventions. Among them, vemurafenib is a selective BRAF inhibitor approved by Food and Drug Administration for clinical practice. Unfortunately, vemurafenib often displays limited efficacy in poorly differentiated and anaplastic thyroid carcinomas probably because of intrinsic and/or acquired resistance mechanisms. In this view, cancer stem cells (CSCs) may represent a possible mechanism of resistance to vemurafenib, due to their self-renewal and chemo resistance properties. RESULTS: We present a computational framework to suggest new potential targets to overcome drug resistance. It has been validated with an in vitro model based upon a spheroid-forming method able to isolate thyroid CSCs that may mimic resistance to vemurafenib. Indeed, vemurafenib did not inhibit cell proliferation of BRAF V600E thyroid CSCs, but rather stimulated cell proliferation along with a paradoxical over-activation of ERK and AKT pathways. The computational model identified a fundamental role of mitogen-activated protein kinase 8 (MAP3K8), a serine/threonine kinase expressed in thyroid CSCs, in mediating this drug resistance. To confirm model prediction, we set a suitable in vitro experiment revealing that the treatment with MAP3K8 inhibitor restored the effect of vemurafenib in terms of both DNA fragmentation and poly (ADP-ribose) polymerase cleavage (apoptosis) in thyroid CSCs. Moreover, MAP3K8 expression levels may be a useful marker to predict the response to vemurafenib. AVAILABILITY AND IMPLEMENTATION: The model is available in GitHub repository visiting the following URL: https://github.com/francescopappalardo/MAP3K8-Thyroid-Spheres-V-3.0. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: DNA methylation is an epigenetic mechanism of genomic regulation involved in the maintenance of homeostatic balance. Dysregulation of DNA methylation status is one of the driver alterations occurring in neoplastic transformation and cancer progression. The identification of methylation hotspots associated to gene dysregulation may contribute to discover new prognostic and diagnostic biomarkers, as well as, new therapeutic targets. RESULTS: We present EpiMethEx (Epigenetic Methylation and Expression), a R package to perform a large-scale integrated analysis by cyclic correlation analyses between methylation and gene expression data. For each gene, samples are segmented according to the expression levels to select genes that are differentially expressed. This stratification allows to identify CG methylation probesets modulated among gene-stratified samples. Subsequently, the methylation probesets are grouped by their relative position in gene sequence to identify wide genomic methylation events statically related to genetic modulation. CONCLUSIONS: The beta-test study showed that the global methylation analysis was in agreement with scientific literature. In particular, this analysis revealed a negative association between promoter hypomethylation and overexpression in a wide number of genes. Less frequently, this overexpression was sustained by intragenic hypermethylation events.

BACKGROUND: In recent years, the study of immune response behaviour using bottom up approach, Agent Based Modeling (ABM), has attracted considerable efforts. The ABM approach is a very common technique in the biological domain due to high demand for a large scale analysis tools for the collection and interpretation of information to solve biological problems. Simulating massive multi-agent systems (i.e. simulations containing a large number of agents/entities) requires major computational effort which is only achievable through the use of parallel computing approaches. RESULTS: This paper explores different approaches to parallelising the key component of biological and immune system models within an ABM model: pairwise interactions. The focus of this paper is on the performance and algorithmic design choices of cell interactions in continuous and discrete space where agents/entities are competing to interact with one another within a parallel environment. CONCLUSIONS: Our performance results demonstrate the applicability of these methods to a broader class of biological systems exhibiting typical cell to cell interactions. The advantage and disadvantage of each implementation is discussed showing each can be used as the basis for developing complete immune system models on parallel hardware.

The 2nd Computational Methods for the Immune System function Workshop has been held in Madrid in conjunction with the IEEE International Conference on Bioinformatics and Biomedicine (BIBM 2018) in Madrid, Spain, from December 3 to 6, 2018. The workshop has been obtained 100% more submissions in respect to the first edition, highlighting a growing interest for the treated topics. The best papers (9) have been selected for extension in this special issue, with themes about immune system and disease simulation, computer-aided design of novel candidate vaccines, methods for the analysis of immune system involved diseases based on statistical methods, meta-heuristics and game theory, and modelling strategies for improving the simulation of the immune system dynamics.

BACKGROUND: Tuberculosis (TB) represents a worldwide cause of mortality (it infects one third of the world's population) affecting mostly developing countries, including India, and recently also developed ones due to the increased mobility of the world population and the evolution of different new bacterial strains capable to provoke multi-drug resistance phenomena. Currently, antitubercular drugs are unable to eradicate subpopulations of Mycobacterium tuberculosis (MTB) bacilli and therapeutic vaccinations have been postulated to overcome some of the critical issues related to the increase of drug-resistant forms and the difficult clinical and public health management of tuberculosis patients. The Horizon 2020 EC funded project "In Silico Trial for Tuberculosis Vaccine Development" (STriTuVaD) to support the identification of new therapeutic interventions against tuberculosis through novel in silico modelling of human immune responses to disease and vaccines, thereby drastically reduce the cost of clinical trials in this critical sector of public healthcare. RESULTS: We present the application of the Universal Immune System Simulator (UISS) computational modeling infrastructure as a disease model for TB. The model is capable to simulate the main features and dynamics of the immune system activities i.e., the artificial immunity induced by RUTI® vaccine, a polyantigenic liposomal therapeutic vaccine made of fragments of Mycobacterium tuberculosis cells (FCMtb). Based on the available data coming from phase II Clinical Trial in subjects with latent tuberculosis infection treated with RUTI® and isoniazid, we generated simulation scenarios through validated data in order to tune UISS accordingly to STriTuVaD objectives. The first case simulates the establishment of MTB latent chronic infection with some typical granuloma formation; the second scenario deals with a reactivation phase during latent chronic infection; the third represents the latent chronic disease infection scenario during RUTI® vaccine administration. CONCLUSIONS: The application of this computational modeling strategy helpfully contributes to simulate those mechanisms involved in the early stages and in the progression of tuberculosis infection and to predict how specific therapeutical strategies will act in this scenario. In view of these results, UISS owns the capacity to open the door for a prompt integration of in silico methods within the pipeline of clinical trials, supporting and guiding the testing of treatments in patients affected by tuberculosis.

BACKGROUND: Multiple Sclerosis (MS) is an immune-mediated inflammatory disease of the Central Nervous System (CNS) which damages the myelin sheath enveloping nerve cells thus causing severe physical disability in patients. Relapsing Remitting Multiple Sclerosis (RRMS) is one of the most common form of MS in adults and is characterized by a series of neurologic symptoms, followed by periods of remission. Recently, many treatments were proposed and studied to contrast the RRMS progression. Among these drugs, daclizumab (commercial name Zinbryta), an antibody tailored against the Interleukin-2 receptor of T cells, exhibited promising results, but its efficacy was accompanied by an increased frequency of serious adverse events. Manifested side effects consisted of infections, encephalitis, and liver damages. Therefore daclizumab has been withdrawn from the market worldwide. Another interesting case of RRMS regards its progression in pregnant women where a smaller incidence of relapses until the delivery has been observed. RESULTS: In this paper we propose a new methodology for studying RRMS, which we implemented in GreatSPN, a state-of-the-art open-source suite for modelling and analyzing complex systems through the Petri Net (PN) formalism. This methodology exploits: (a) an extended Colored PN formalism to provide a compact graphical description of the system and to automatically derive a set of ODEs encoding the system dynamics and (b) the Latin Hypercube Sampling with PRCC index to calibrate ODE parameters for reproducing the real behaviours in healthy and MS subjects.To show the effectiveness of such methodology a model of RRMS has been constructed and studied. Two different scenarios of RRMS were thus considered. In the former scenario the effect of the daclizumab administration is investigated, while in the latter one RRMS was studied in pregnant women. CONCLUSIONS: We propose a new computational methodology to study RRMS disease. Moreover, we show that model generated and calibrated according to this methodology is able to reproduce the expected behaviours.

The developing of novel prophylactic and therapeutic vaccine candidates in the field of cancer immunology brought to very promising results against tumors, entitling full protection with reduced amount of the typical side effects of the actual conventional treatments. However, such treatments required a constant, life-long, administration procedure to keep protection. As both the period of protection and the relative number of administrations grow, the problem of finding the best administration protocol, in time and dosage, becomes more and more complex. Such a problem cannot be usually solved in in vivo experiments, as the costs in terms of time, money, and people would be prohibitive. We propose a hybrid approach that integrates machine learning and parallel genetic algorithms to enhance the research in silico of optimal administration protocols for a cancer vaccine. A neural network is used to improve both crossover and mutation operators. Preliminary results suggest that the use of such could bring to better administration protocols using a similar computational effort.

Personalized target therapies represent one of the possible treatment strategies to fight the ongoing battle against cancer. New treatment interventions are still needed for an effective and successful cancer therapy. In this scenario, we simulated and analyzed the dynamics of BRAF V600E melanoma patients treated with BRAF inhibitors in order to find potentially interesting targets that may make standard treatments more effective in particularly aggressive tumors that may not respond to selective inhibitor drugs. To this aim, we developed a continuous Petri Net model that simulates fundamental signalling cascades involved in melanoma development, such as MAPK and PI3K/AKT, in order to deeply analyze these complex kinase cascades and predict new crucial nodes involved in melanomagenesis. The model pointed out that some microRNAs, like hsa-mir-132, downregulates expression levels of p120RasGAP: under high concentrations of p120RasGAP, MAPK pathway activation is significantly decreased and consequently also PI3K/PDK1/AKT activation. Furthermore, our analysis carried out through the Genomic Data Commons (GDC) Data Portal shows the evidence that hsa-mir-132 is significantly associated with clinical outcome in melanoma cancer genomic data sets of BRAF-mutated patients. In conclusion, targeting miRNAs through antisense oligonucleotides technology may suggest the way to enhance the action of BRAF-inhibitors.

The aim of this paper is (i)to study breast cancer growth by mean of a mathematical model describing cell population dynamics during cancer growth, and (ii)to use this model to reproduce and explain experimental data. We started from a linear model describing cancer subpopulations evolution based on the Cancer Stem Cell (CSC) theory, and we added feedback mechanisms from the cell populations to mimic micro-environment effects in cancer growth. In details, we hypothesized two feedback mechanisms and we studied their effects both separately and combined together. In this way we obtained three new models that we tuned using data derived by TUBO Cancer cell line and describing the evolution of the total cell population and the subpopulations over time. Finally, we exploited these three models to understand which combination of feedback mechanisms better describe the experimental data.

Vaccines represent nowadays one of the most efficient weapons against foreign pathogens. To be effective, vaccines need a proper administration strategy that requires multiple administrations in order to ensure the acquisition of immunological memory. Vaccination schedules are usually based on past experience, and economical, ethical and time constraints have limited the research for better combinations of timing and dosage. We present here a computational approach based on the use of stochastic optimization techniques and validated “in silico” models to study and optimize vaccination protocols. We use Simulated Annealing in conjunction with a validated agent based model to suggest some interventions that may improve the efficacy of a candidate vaccine composed by influenza-A virosome and natural citrus-derived adjuvants to prevent influenza-A infection.

Immune system processes can be simulated using both system dynamics (SD) models based on differential equations and Agent-based (AB) models. The two approaches are intrinsically different but some methodologies have been developed to convert SD models into AB models with a variable degree of success. However, until now none of such methods have considered the use of scale factors in SD to AB model conversion. In this work, we revisited a well know SD model describing the interaction between effector T cells and tumor cells that was previously shown unsuitable for AB modeling. We introduced non-dimensional scaling factors in AB modeling and compared AB and AD simulations through a sensitivity analysis. Under this scenario, we obtained AB models that could successfully reproduce SD simulations with a reasonable number of agents and a stochastic behavior that did not compromise computer resources. In general, our results justify the introduction of non-dimensional scaling factors to reproduce SD simulations with AB models.

BACKGROUND: Human papillomavirus infection is a global social burden that, every year, leads to thousands new diagnosis of cancer. The introduction of a protocol of immunization, with Gardasil and Cervarix vaccines, has radically changed the way this infection easily spreads among people. Even though vaccination is only preventive and not therapeutic, it is a strong tool capable to avoid the consequences that this pathogen could cause. Gardasil vaccine is not free from side effects and the duration of immunity is not always well determined. This work aim to enhance the effects of the vaccination by using a new class of adjuvants and a different administration protocol. Due to their minimum side effects, their easy extraction, their low production costs and their proven immune stimulating activity, citrus-derived molecules are valid candidates to be administered as adjuvants in a vaccine formulation against Hpv. RESULTS: With the aim to get a stronger immune response against Hpv infection we built an in silico model that delivers a way to predict the best adjuvants and the optimal means of administration to obtain such a goal. Simulations envisaged that the use of Neohesperidin elicited a strong immune response that was then validated in vivo. CONCLUSIONS: We built up a computational infrastructure made by a virtual screening approach able to preselect promising citrus derived compounds, and by an agent based model that reproduces HPV dynamics subject to vaccine stimulation. This integrated methodology was able to predict the best protocol that confers a very good immune response against HPV infection. We finally tested the in silico results through in vivo experiments on mice, finding good agreement.

MOTIVATION: Vaccines represent the most effective and cost-efficient weapons against a wide range of diseases. Nowadays new generation vaccines based on subunit antigens reduce adverse effects in high risk individuals. However, vaccine antigens are often poor immunogens when administered alone. Adjuvants represent a good strategy to overcome such hurdles, indeed they are able to: enhance the immune response; allow antigens sparing; accelerate the specific immune response; and increase vaccine efficacy in vulnerable groups such as newborns, elderly or immuno-compromised people. However, due to safety concerns and adverse reactions, there are only a few adjuvants approved for use in humans. Moreover, in practice current adjuvants sometimes fail to confer adequate stimulation. Hence, there is an imperative need to develop novel adjuvants that overcome the limitations of the currently available licensed adjuvants. RESULTS: We developed a computational framework that provides a complete pipeline capable of predicting the best citrus-derived adjuvants for enhancing the immune system response using, as a target disease model, influenza A infection. In silico simulations suggested a good immune efficacy of specific citrus-derived adjuvant (Beta Sitosterol) that was then confirmed in vivoAvailability: The model is available visiting the following URL: http://vaima.dmi.unict.it/AdjSim CONTACT: [email protected]; [email protected].

BACKGROUND: Mathematical and computational models showed to be a very important support tool for the comprehension of the immune system response against pathogens. Models and simulations allowed to study the immune system behavior, to test biological hypotheses about diseases and infection dynamics, and to improve and optimize novel and existing drugs and vaccines. Continuous models, mainly based on differential equations, usually allow to qualitatively study the system but lack in description; conversely discrete models, such as agent based models and cellular automata, permit to describe in detail entities properties at the cost of losing most qualitative analyses. Petri Nets (PN) are a graphical modeling tool developed to model concurrency and synchronization in distributed systems. Their use has become increasingly marked also thanks to the introduction in the years of many features and extensions which lead to the born of "high level" PN. RESULTS: We propose a novel methodological approach that is based on high level PN, and in particular on Colored Petri Nets (CPN), that can be used to model the immune system response at the cellular scale. To demonstrate the potentiality of the approach we provide a simple model of the humoral immune system response that is able of reproducing some of the most complex well-known features of the adaptive response like memory and specificity features. CONCLUSIONS: The methodology we present has advantages of both the two classical approaches based on continuous and discrete models, since it allows to gain good level of granularity in the description of cells behavior without losing the possibility of having a qualitative analysis. Furthermore, the presented methodology based on CPN allows the adoption of the same graphical modeling technique well known to life scientists that use PN for the modeling of signaling pathways. Finally, such an approach may open the floodgates to the realization of multi scale models that integrate both signaling pathways (intra cellular) models and cellular (population) models built upon the same technique and software.

MOTIVATION: Many important problems in cell biology require dense non-linear interactions between functional modules to be considered. The importance of computer simulation in understanding cellular processes is now widely accepted, and a variety of simulation algorithms useful for studying certain subsystems have been designed. Expansion of hematopoietic stem and progenitor cells (HSC/HPC) in ex vivo culture with cytokines and small molecules is a method to increase the restricted numbers of stem cells found in umbilical cord blood (CB), while also enhancing the content of early engrafting neutrophil and platelet precursors. The efficacy of the expanded product depends on the composition of the cocktail of cytokines and small molecules used for culture. Testing the influence of a cytokine or small molecule on the expansion of HSC/HPC is a laborious and expensive process. We therefore developed a computational model based on cellular signaling interactions that predict the influence of a cytokine on the survival, duplication and differentiation of the CD133(+) HSC/HPC subset from human umbilical CB. RESULTS: We have used results from in vitro expansion cultures with different combinations of one or more cytokines to develop an ordinary differential equation model that includes the effect of cytokines on survival, duplication and differentiation of the CD133(+) HSC/HPC. Comparing the results of in vitro and in silico experiments, we show that the model can predict the effect of a cytokine on the fold expansion and differentiation of CB CD133(+) HSC/HPC after 8-day culture on a 3D scaffold. Supplementary data are available at Bioinformatics online.

MOTIVATION: Although results from phase III clinical trials substantially support the use of prophylactic and therapeutic vaccines against cancer, what has yet to be defined is how many and how frequent boosts are needed to sustain a long-lasting and protecting memory T-cell response against tumor antigens. Common experience is that such preclinical tests require the sacrifice of a relatively large number of animals, and are particularly time- and money-consuming. RESULTS: As a first step to overcome these hurdles, we have developed an ordinary differential equation model that includes all relevant entities (such as activated cytotoxic T lymphocytes and memory T cells), and investigated the induction of immunological memory in the context of wild-type mice injected with a dendritic cell-based vaccine. We have simulated the biological behavior both in the presence and in the absence of memory T cells. Comparing results of ex vivo and in silico experiments, we show that the model is able to envisage the expansion and persistence of antigen-specific memory T cells. The model might be applicable to more complex vaccination schedules and substantially in any biological condition of prime-boosting. AVAILABILITY AND IMPLEMENTATION: The model is fully described in the article.

BACKGROUND: Multiple sclerosis (MS) is a disease of central nervous system that causes the removal of fatty myelin sheath from axons of the brain and spinal cord. Autoimmunity plays an important role in this pathology outcome and body's own immune system attacks on the myelin sheath causing the damage. The etiology of the disease is partially understood and the response to treatment cannot easily be predicted. RESULTS: We presented the results obtained using 8 genetically predisposed randomly chosen individuals reproducing both the absence and presence of malfunctions of the Teff-Treg cross-balancing mechanisms at a local level. For simulating the absence of a local malfunction we supposed that both Teff and Treg populations had similar maximum duplication rates. Results presented here suggest that presence of a genetic predisposition is not always a sufficient condition for developing the disease. Other conditions such as a breakdown of the mechanisms that regulate and allow peripheral tolerance should be involved. CONCLUSIONS: The presented model allows to capture the essential dynamics of relapsing-remitting MS despite its simplicity. It gave useful insights that support the hypothesis of a breakdown of Teff-Treg cross balancing mechanisms.

The definition of artificial immunity, realized through vaccinations, is nowadays a practice widely developed in order to eliminate cancer disease. The present paper deals with an improved version of a mathematical model recently analyzed and related to the competition between immune system cells and mammary carcinoma cells under the action of a vaccine (Triplex). The model describes in detail both the humoral and cellular response of the immune system to the tumor associate antigen and the recognition process between B cells, T cells and antigen presenting cells. The control of the tumor cells growth occurs through the definition of different vaccine protocols. The performed numerical simulations of the model are in agreement with in vivo experiments on transgenic mice.

BACKGROUND: The Triplex cell vaccine is a cancer cellular vaccine that can prevent almost completely the mammary tumor onset in HER-2/neu transgenic mice. In a translational perspective, the activity of the Triplex vaccine was also investigated against lung metastases showing that the vaccine is an effective treatment also for the cure of metastases. A future human application of the Triplex vaccine should take into account several aspects of biological behavior of the involved entities to improve the efficacy of therapeutic treatment and to try to predict, for example, the outcomes of longer experiments in order to move faster towards clinical phase I trials. To help to address this problem, MetastaSim, a hybrid Agent Based - ODE model for the simulation of the vaccine-elicited immune system response against lung metastases in mice is presented. The model is used as in silico wet-lab. As a first application MetastaSim is used to find protocols capable of maximizing the total number of prevented metastases, minimizing the number of vaccine administrations. RESULTS: The model shows that it is possible to obtain "in silico" a 45% reduction in the number of vaccinations. The analysis of the results further suggests that any optimal protocol for preventing lung metastases formation should be composed by an initial massive vaccine dosage followed by few vaccine recalls. CONCLUSIONS: Such a reduction may represent an important result from the point of view of translational medicine to humans, since a downsizing of the number of vaccinations is usually advisable in order to minimize undesirable effects. The suggested vaccination strategy also represents a notable outcome. Even if this strategy is commonly used for many infectious diseases such as tetanus and hepatitis-B, it can be in fact considered as a relevant result in the field of cancer-vaccines immunotherapy. These results can be then used and verified in future "in vivo" experiments, and their outcome can be used to further improve and refine the model.

Vaccine research is a combinatorial science requiring computational analysis of vaccine components, formulations and optimization. We have developed a framework that combines computational tools for the study of immune function and vaccine development. This framework, named ImmunoGrid combines conceptual models of the immune system, models of antigen processing and presentation, system-level models of the immune system, Grid computing, and database technology to facilitate discovery, formulation and optimization of vaccines. ImmunoGrid modules share common conceptual models and ontologies. The ImmunoGrid portal offers access to educational simulators where previously defined cases can be displayed, and to research simulators that allow the development of new, or tuning of existing, computational models. The portal is accessible at .

SUMMARY: Since few years the problem of finding optimal solutions for drug or vaccine protocols have been tackled using system biology modeling. These approaches are usually computationally expensive. Our previous experiences in optimizing vaccine or drug protocols using genetic algorithms required the use of a high performance computing infrastructure for a couple of days. In the present article we show that by an appropriate use of a different optimization algorithm, the simulated annealing, we have been able to downsize the computational effort by a factor 10(2). The new algorithm requires computational effort that can be achieved by current generation personal computers. AVAILABILITY: Software and additional data can be found at http://www.immunomics.eu/SA/