Network-based approaches to study human health and disease
Feb 18, 2025
Andreas Zanzoni is an Associate Professor of Bioinformatics and Genomics at the Department of Biology, Aix-Marseille University. His research focuses on understanding the molecular basis of human diseases through network-based approaches, with a particular emphasis on:
• Host-microbe protein interactions in chronic diseases
• The role of protein-RNA interactions in cellular networks
• Multifunctional proteins, regulatory sites, and signal transduction
• Mapping host-pathogen interactomes, particularly between coronaviruses and the human host
|
How artificial intelligence is reshaping computer-aided drug design
June 94, 2025
Computer-aided drug design (CADD) has become a key tool in modern drug discovery, significantly accelerating development while reducing costs. Recent advancements in CADD integrate various computational approaches, combining physics-based modeling with emerging artificial intelligence (AI) technologies. I will illustrate how atomic-level simulations have played a key role in identifying new drug candidates for diseases with major socio-economic impact, while advancements in AI, driven by increasing data availability and continuous model refinement, are unlocking new possibilities for predicting therapeutic target structures, analyzing molecular properties, and designing novel molecules.
|
Comprehensive in silico characterization of mitochondrial carriers of the SLC25 protein family
June 11, 2025
Mitochondria and cytosol metabolic processes require the import/export of metabolites across the impermeable inner mitochondrial membrane. The transport of these molecules is performed by a family of membrane transporters known as the mitochondrial carriers (MCs), which act as master regulators of cell metabolism in health and disease. Further, MCs are key players in the reprogramming of cell metabolism that occurs in cancer. Several of them are overexpressed in various cancer types as essential gatekeepers of the trafficking of metabolic intermediates. These represent promising targets to develop novel therapies aimed at restoring the physiological metabolic processes, reducing cancer cell proliferation and metastasis formation.
In this framework, in the last few years we employed a computation approach to uncover the structural determinants of endogenous and exogenous (drug-like) ligands recognition of several members of this protein family. In addition, we also developed a computational protocol for the de-orphanization of transporters whose ligand is still unknown with the aim of uncovering novel mitochondrial metabolic pathways. Prominent results of these activities will be illustrated and discussed.
|
From Generative AI Foundations to Protein Language Models
July 10, 2025
This talk will trace the evolution of modern generative AI—from the core principles and architectures that power large-scale language models to their adaptation for understanding and designing proteins. You’ll learn how techniques like self‑supervised learning, attention mechanisms, and transformer scaling have been repurposed to capture the “language” of amino acid sequences, enabling breakthroughs in structure prediction, functional annotation, and de novo protein design. By the end, you’ll see how foundational AI research is unlocking entirely new frontiers in biology and bioengineering.
|
From Protein Structure to Mechanism: Integrating Modelling, Simulations, and AI in the Study of Rare Diseases
Nov 24, 2025
My research journey has focused on understanding how proteins function at the molecular level through computational modelling and simulations. Trained as a structural bioinformatician, I have explored the conformational dynamics and substrate binding properties of membrane proteins. Over time, I became increasingly interested in the potential of artificial intelligence to enhance and accelerate these studies. By combining physics-based simulations with AI-driven approaches, I aim to bridge the gap between molecular-scale insight and biological complexity. At TIGEM, my work now focuses on applying these computational strategies to the study of rare diseases, where detailed molecular understanding can illuminate pathogenic mechanisms and guide therapeutic discovery. Indeed, our main project involves the computational design of protein variants with enhanced stability and/or activity that can be used to improve enzyme replacement and gene therapies.
|
From Big Omics Data to Structural Modeling Supported by High-Performance Computing
Jun 23rd, 2026
This seminar offers an overview of various applications of bioinformatics to ecology and molecular modeling, demonstrating how complex biological data can be transformed into useful resources, interpretable results, and new scientific hypotheses. Through examples drawn from recent research projects, approaches for analyzing omics data will be discussed, with a particular focus on transcriptomics, comparative genomics, and the functional interpretation of data from non-model organisms.
Furthermore, examples of structural modeling, docking, and molecular dynamics applied to the study of molecular interactions and potential mechanisms of action will be presented. High-performance computing (HPC) will be highlighted as a supporting tool to tackle large-scale analyses, automate complex workflows, and streamline the processing of big biological data.
Among the research lines, the seminar will also illustrate:
An innovative pipeline based on Large Language Models (LLMs) for automated knowledge extraction from prebiotic chemistry scientific literature.
The development of a computational framework for automating Umbrella Sampling workflows, aimed at calculating ligand-receptor binding free energy on HPC architectures.
A prototype genome assembly algorithm implemented on quantum computing platforms.
|
seminar
|