Mechanistic-based Model simulations (MM) are an effective approach commonly employed, for research and learning purposes, to better investigate and understand the inherent behavior of biological systems. Recent advancements in modern technologies and the large availability of omics data allowed the application of Machine Learning (ML) techniques to different research fields, including systems biology. However, the availability of information regarding the analyzed biological context, sufficient experimental data, as well as the degree of computational complexity, represent some of the issues that both MMs and ML techniques could present individually. For this reason, recently, several studies suggest overcoming or significantly reducing these drawbacks by combining the above-mentioned two methods. In the wake of the growing interest in this hybrid analysis approach, with the present review, we want to systematically investigate the studies available in the scientific literature in which both MMs and ML have been combined to explain biological processes at genomics, proteomics, and metabolomics levels, or the behavior of entire cellular populations. Elsevier Scopus®, Clarivate Web of Science™ and National Library of Medicine PubMed® databases were enquired using the queries reported in Table 1, resulting in 350 scientific articles. Only 14 of the 350 documents returned by the comprehensive search conducted on the three major online databases met our search criteria, i.e. present a hybrid approach consisting of the synergistic combination of MMs and ML to treat a particular aspect of systems biology. Despite the recent interest in this methodology, from a careful analysis of the selected papers, it emerged how examples of integration between MMs and ML are already present in systems biology, highlighting the great potential of this hybrid approach to both at micro and macro biological scales.

The study of novel drug delivery systems represents one of the frontiers of the biomedical research area. Multi-disciplinary scientific approaches combining traditional or engineered technologies are used to provide major advances in improving drug bioavailability, rate of release, cell/tissue specificity and therapeutic index. Biodegradable and bio-absorbable polymers are usually the building blocks of these systems, and their copolymers are employed to create delivery components. For example, poly (lactic acid) or poly (glycolic acid) are often used as bricks for the production drug-based delivery systems as polymeric microparticles (MPs) or micron-scale needles. To avoid time-consuming empirical approaches for the optimization of these formulations, in silico-supported models have been developed. These methods can predict and tune the release of different drugs starting from designed combinations. Starting from these considerations, this review has the aim of investigating recent approaches to the production of polymeric carriers and the combination of in silico and experimental methods as promising platforms in the biomedical field.

Until the approval of vaccines at the end of 2020, societies relied on non-pharmaceutical interventions (NPIs) in order to control the COVID-19 pandemic. Spontaneous changes in individual behavior might have contributed to or counteracted epidemic control due to NPIs. For example, the population compliance to NPIs may have varied over time as people developed “epidemic fatigue” or altered their perception of the risk and severity of COVID-19. Whereas official measures are well documented, the behavioral response of the citizens is harder to capture. We propose a mathematical model of the societal response, taking into account three main effects: the citizen response dynamics, the authorities’ NPIs, and the occurrence of unpreventable events that significantly alter the virus transmission rate. A key assumption is that a society has a waning memory of the epidemic effects, which reflects on both the severity of the authorities’ NPIs and on the citizens’ compliance to the prescribed rules. This, in turn, feeds back onto the transmission rate of the disease, such that a higher number of hospitalizations decreases the probability of transmission. We show that the model is able to reproduce the COVID-19 dynamics in terms of hospital admissions for several European countries during 2020 over surprisingly long time scales. Also, it is capable of capturing the effects of disturbances (for example the emergence of new virus variants) and can be exploited for implementing control actions to limit such effects. A possible application, illustrated in this letter, consists of exploiting the estimations based on the data of one country, to predict and control the evolution in another country, where the virus spreading is still in an earlier phase.

Arrhythmogenic Right Ventricular cardiomyopathy (ARVC) is an inherited cardiac muscle disease linked to genetic deficiency in components of the desmosomes. The disease is characterized by progressive fibro-fatty replacement of the right ventricle, which acts as a substrate for arrhythmias and sudden cardiac death. The molecular mechanisms underpinning ARVC are largely unknown. Here we propose a mathematical model for investigating the molecular dynamics underlying heart remodeling and the loss of cardiac myocytes identity during ARVC. Our methodology is based on three computational models: firstly, in the context of the Wnt pathway, we examined two different competition mechanisms between β-catenin and Plakoglobin (PG) and their role in the expression of adipogenic program. Secondly, we investigated the role of RhoA-ROCK pathway in ARVC pathogenesis, and thirdly we analyzed the interplay between Wnt and RhoA-ROCK pathways in the context of the ARVC phenotype. We conclude with the following remark: both Wnt/β-catenin and RhoA-ROCK pathways must be inactive for a significant increase of PPARγ expression, suggesting that a crosstalk mechanism might be responsible for mediating ARVC pathogenesis.

The study of homeostatic equilibrium is a key concern in several fields, from physiology and biology to medicine and biomedical engineering. Control theory approaches can provide effective strategies to model physiological control systems, helping in understanding the dynamics of bio- and physio-logical regulation processes. However, the intrinsic complexity of living systems makes it difficult to identify unified biomodels that can represent a wide variety of physiological systems. In this context, the present work proposes a general framework to model the dynamics and describe the behavior of a wide class of multivariable physiological control systems, from the molecular to the whole-organ scale. The framework adopts a structure based on a closed-loop topology taking into account multiple inputs and outputs and with the negative feedback action intrinsically embedded within the model. The development of such a general model has at least three important repercussions: the first concerns the possibility of better understanding the basic mechanisms common to many physiological systems; the second is to develop a common theoretical framework to enable effective approaches to the analysis and design of synthetic biological control systems; finally, the investigation of the structural properties of the model in a general context, allows a guided and simplified application to specific cases. To this regard, in this paper, the existence, possible uniqueness and stability properties of the homeostatic equilibrium points of the general model are investigated; the theoretical framework is then illustrated through two real-world case-studies: (i) the PI3K/AKT/mTOR pathway nonlinear dynamics, a critical regulator of cellular growth, proliferation, and survival; (ii) the control mechanism of the neuromuscular stretch reflex, among the prime triggers implicated in postural control. Results proved the capability of the proposed framework to capture the intricate dynamics of multivariable physiological systems at different scales, highlighting the existence of asymptotically stable homeostatic equilibrium and allowing the study of the impact of transmission delays on the system’s stability. At the best of authors’ knowledge, following the paper [1] where monovariable systems where dealt with, the proposed methodology is the first attempt to represent and investigate homeostasis from the molecular up to systemic level by exploiting a unified multivariable biomodeling architecture, which makes it a novel approach to understanding homeostatic control from a broader perspective.

Binge eating disorder (BED) carries a 6 times higher risk for obesity and accounts for roughly 30% of type 2 diabetes cases. Timely identification of early glycemic disturbances and comprehensive treatment can impact on the likelihood of associated metabolic complications and the overall outcome. In this study, machine learning techniques were applied to static and dynamic glucose-derived measures to detect BED among 281 individuals with high weight. Data from the classic (2 h) and the extended (5 h) glucose load were computed by multiple algorithms and two models with the most relevant features were trained to detect BED within the sample. The models were then tested on an independent cohort (N = 21). The model based on the 5 h-long glucose load exhibited the best performance (sensitivity = 0.75, specificity = 0.67, F score = 0.71) diagnosing BED in 7 out of 10 cases. Sex, HOMA-IR, HbA1c and plasma glucose in different times, and hypoglycemia events were the most sensitive features for BED diagnosis. This study is the first to use metabolic hallmarks to train ML algorithms for detecting BED in individuals at high risk for metabolic complications. ML techniques applied to objective and reliable glycemic features might prompt the identification of BED among individuals at high risk for metabolic complications, enabling timely and tailored multidisciplinary treatment.