arXiv subjects

Ion Petre

Publications and source records attributed to Ion Petre.

13 recordsLinked to original sources

arXiv2026

A Unified Framework for Reaction Systems Based on Interval Structures

Reaction systems have evolved into a rich family of computational models differing in their treatment of multiplicities, resource management, concurrency, and state evolution. We introduce a unified semantic framework based on interval structures and interval-based transformation systems. The framework decomposes operational semantics into independent resource, production, update, and execution strategies, providing a common basis for describing, comparing, and constructing reaction-system variants. We show that classical reaction systems, restricted reaction systems, multiset reaction systems, reaction systems with concentration, and resource-preserving multiset reaction systems are all recovered as instantiations of the framework. Quantitative reaction systems are accommodated through an additional preprocessing stage. We further demonstrate that the framework naturally extends beyond reaction systems to other computational models, including Petri nets. The proposed framework provides a common semantic foundation for existing models and a flexible basis for developing and analysing new computational formalisms.

cs.LO↗

arXiv2026

Generative Augmentation of Raman Spectra for Glioma Classification

Access to sufficiently large biomedical datasets remains a major obstacle for machine learning in Raman spectroscopy-based diagnostics. In particular, for glioma analysis, datasets are typically small and heterogeneous, affected by acquisition-specific variability. This work investigates the utility of deep generative augmentation in such a small-cohort setting. We analyze glioma biopsy spectra acquired from 58 tumor samples and consider both binary IDH-status classification and 6-class methylation subtype classification problems. To address the limited size and imbalance of the dataset, we develop a conditional variational autoencoder ($β$-CVAE) capable of generating class-conditioned synthetic Raman spectra. The generated data are evaluated in Train-on-Synthetic, Test-on-Real (TS/TR) and Train-on-Synthetic+Real, Test-on-Real (TSR/TR) settings under a strict patient-isolated cross-validation protocol. Models trained exclusively on synthetic data underperform models trained on real spectra, indicating a substantial domain gap between synthetic and real distributions. However, augmenting the real training data with synthetic spectra consistently improves classification performance across multiple models. These findings indicate that, even with a limited number of independent patient samples, generative models can capture sufficient structure to provide useful regularization for downstream classifiers. We also investigate a reconstruction-based inference strategy, termed Classification by Reconstruction (CbR), in which class prediction is based on reconstruction error under different class conditions. Overall, the results support the use of deep generative augmentation as a practical strategy for improving machine learning robustness in Raman spectroscopy applications characterized by limited biomedical datasets.

cs.LG↗

arXiv2026

Machine Learning for Biomedical Raman Spectroscopy: From Spectral Acquisition to Clinical Translation

Raman spectroscopy provides label-free, chemically specific characterization of biological systems and has become an important tool for cancer diagnosis, molecular subtyping, microbiological identification, and intraoperative decision support. Biomedical Raman spectra are, however, high-dimensional, noisy, and affected by fluorescence background, acquisition variability, and biological heterogeneity, making robust computational analysis essential. This review examines the role of machine learning across the biomedical Raman spectroscopy pipeline, from preprocessing and signal correction to unsupervised structure discovery, supervised diagnosis and molecular stratification, representation and transfer learning, explainability, biomarker discovery, and multimodal integration with imaging, pathology, and molecular profiling. Emphasis is placed on the use of machine learning not only for diagnostic classification, but also for biologically interpretable and clinically actionable analysis. We also discuss the main barriers to clinical translation, including limited dataset sizes, inter-instrument variability, inconsistent preprocessing, insufficient external validation, reproducibility concerns, and limited sharing of software, data, and metadata. We argue that progress will require methodological advances together with standardization, robust validation, explainability, and deployment-ready analytical frameworks. By integrating methodological, biomedical, and translational perspectives, this review outlines key directions for developing reliable and clinically deployable Raman-AI systems.

cs.LG↗

arXiv2024

Strong regulatory graphs

Logical modeling is a powerful tool in biology, offering a system-level understanding of the complex interactions that govern biological processes. A gap that hinders the scalability of logical models is the need to specify the update function of every vertex in the network depending on the status of its predecessors. To address this, we introduce in this paper the concept of strong regulation, where a vertex is only updated to active/inactive if all its predecessors agree in their influences; otherwise, it is set to ambiguous. We explore the interplay between active, inactive, and ambiguous influences in a network. We discuss the existence of phenotype attractors in such networks, where the status of some of the variables is fixed to active/inactive, while the others can have an arbitrary status, including ambiguous.

cs.DM↗

arXiv2021

Network analytics for drug repurposing in COVID-19

To better understand the potential of drug repurposing in COVID-19, we analyzed control strategies over essential host factors for SARS-CoV-2 infection. We constructed comprehensive directed protein-protein interaction networks integrating the top ranked host factors, drug target proteins, and directed protein-protein interaction data. We analyzed the networks to identify drug targets and combinations thereof that offer efficient control over the host factors. We validated our findings against clinical studies data and bioinformatics studies. Our method offers a new insight into the molecular details of the disease and into potentially new therapy targets for it. Our approach for drug repurposing is significant beyond COVID-19 and may be applied also to other diseases.

q-bio.MN↗

arXiv2021

Towards Scalable Modeling of Biology in Event-B

Biology offers many examples of large-scale, complex, concurrent systems: many processes take place in parallel, compete on resources and influence each other's behavior. The scalable modeling of biological systems continues to be a very active field of research. In this paper we introduce a new approach based on Event-B, a state-based formal method with refinement as its central ingredient, allowing us to check for model consistency step-by-step in an automated way. Our approach based on functions leads to an elegant and concise modeling method. We demonstrate this approach by constructing what is, to our knowledge, the largest ever built Event-B model, describing the ErbB signaling pathway, a key evolutionary pathway with a significant role in development and in many types of cancer. The Event-B model for the ErbB pathway describes 1320 molecular reactions through 242 events.

q-bio.MN↗

arXiv2021

Network modeling methods for precision medicine

We discuss in this survey several network modeling methods and their applicability to precision medicine. We review several network centrality methods (degree centrality, closeness centrality, eccentricity centrality, betweenness centrality, and eigenvector-based prestige) and two systems controllability methods (minimum dominating sets and network structural controllability). We demonstrate their applicability to precision medicine on three multiple myeloma patient disease networks. Each network consists of protein-protein interactions built around a specific patient's mutated genes, around the targets of the drugs used in the standard of care in multiple myeloma, and around multiple myeloma-specific essential genes. For each network we demonstrate how the network methods we discuss can be used to identify personalized, targeted drug combinations uniquely suited to that patient.

q-bio.MN↗

arXiv2020

Controllability of reaction systems

Controlling a dynamical system is the ability of changing its configuration arbitrarily through a suitable choice of inputs. It is a very well studied concept in control theory, with wide ranging applications in medicine, biology, social sciences, engineering. We introduce in this article the concept of controllability of reaction systems as the ability of transitioning between any two states through a suitable choice of context sequences. We show that the problem is PSPACE-hard. We also introduce a model of oncogenic signalling based on reaction systems and use it to illustrate the intricacies of the controllability of reaction systems.

eess.SY↗

arXiv2020

Identifying efficient controls of complex interaction networks using genetic algorithms

Control theory has seen recently impactful applications in network science, especially in connections with applications in network medicine. A key topic of research is that of finding minimal external interventions that offer control over the dynamics of a given network, a problem known as network controllability. We propose in this article a new solution for this problem based on genetic algorithms. We tailor our solution for applications in computational drug repurposing, seeking to maximise its use of FDA-approved drug targets in a given disease-specific protein-protein interaction network. We show how our algorithm identifies a number of potentially efficient drugs for breast, ovarian, and pancreatic cancer. We demonstrate our algorithm on several benchmark networks from cancer medicine, social networks, electronic circuits, and several random networks with their edges distributed according to the Erdős-Rényi, the small-world, and the scale-free properties. Overall, we show that our new algorithm is more efficient in identifying relevant drug targets in a disease network, advancing the computational solutions needed for new therapeutic and drug repurposing approaches.

q-bio.MN↗

arXiv2013

Proceedings Fourth International Workshop on Computational Models for Cell Processes

The fourth international workshop on Computational Models for Cell Processes (CompMod 2013) took place on June 11, 2013 at the Åbo Akademi University, Turku, Finland, in conjunction with iFM 2013. The first edition of the workshop (2008) took place in Turku, Finland, in conjunction with Formal Methods 2008, the second edition (2009) took place in Eindhoven, the Netherlands, as well in conjunction with Formal Methods 2009, and the third one took place in Aachen, Germany, in conjunction with CONCUR 2013. This volume contains the final versions of all contributions accepted for presentation at the workshop. The goal of the CompMod workshop series is to bring together researchers in Computer Science and Mathematics (both discrete and continuous), interested in the opportunities and the challenges of Systems Biology. The Program Committee of CompMod 2013 selected 3 papers for presentation at the workshop. In addition, we had two invited talks and five informal presentations. The scientific program of the workshop spans an interesting mix of approaches to systems and even synthetic biology, encompassing several different modeling approaches, ranging from quantitative to qualitative techniques, from continuous to discrete mathematics, and from deterministic to stochastic methods. We thank our invited speakers Daniela Besozzi (Universita degli Studi di Milano, Milano, Italy) and Juho Rousu (Aalto University, Finland) for accepting our invitation and for presenting some of their recent results at CompMod 2013. The technical contributions address the mathematical modeling of the PDGF signalling pathway, the canonical labelling of site graphs, rule-based modeling of polymerization reactions, rule-based modeling as a platform for the analysis of synthetic self-assembled nano-systems, robustness analysis of stochastic systems, an algebraic approach to gene assembly in ciliates, and large-scale text mining of biomedical literature.

cs.CE↗

arXiv2011

Proceedings Third International Workshop on Computational Models for Cell Processes

This volume contains the final versions of the papers presented at the 3rd International Workshop on Computational Models for Cell Processes (CompMod 2011). The workshop took place on September 10, 2011 at the University of Aachen, Germany, in conjunction with CONCUR 2011. The first edition of the workshop (2008) took place in Turku, Finland, in conjunction with Formal Methods 2008 and the second edition (2009) took place in Eindhoven, the Netherlands, as well in conjunction with Formal Methods 2009. The goal of the CompMod workshop series is to bring together researchers in Computer Science (especially in Formal Methods) and Mathematics (both discrete and continuous), interested in the opportunities and the challenges of Systems Biology.

cs.CE↗

arXiv2010

Matrix Insertion-Deletion Systems

In this article, we consider for the first time the operations of insertion and deletion working in a matrix controlled manner. We show that, similarly as in the case of context-free productions, the computational power is strictly increased when using a matrix control: computational completeness can be obtained by systems with insertion or deletion rules involving at most two symbols in a contextual or in a context-free manner and using only binary matrices.

cs.FL↗

arXiv2009

Proceedings Second International Workshop on Computational Models for Cell Processes

The second international workshop on Computational Models for Cell Processes (ComProc 2009) took place on November 3, 2009 at the Eindhoven University of Technology, in conjunction with Formal Methods 2009. The workshop was jointly organized with the EC-MOAN project. This volume contains the final versions of all contributions accepted for presentation at the workshop.

cs.CE↗