Date: 3 November 2026

Time: 15:00 CET

Abstract

The BioExcel Building Blocks (BioBB) library was launched in 2019 with the goal of making biomolecular simulation workflows more FAIR (Findable, Accessible, Interoperable, and Reusable/Reproducible). A key objective of the initiative has been to promote interoperability among biomolecular simulation tools while enhancing the reproducibility, portability, and reusability of computational workflows. Since its inception, the library has continuously evolved, expanding its capabilities and supporting a growing range of simulation and analysis applications. This webinar will present the major developments introduced in recent years, including newly integrated software packages, enhanced installation and execution options, and novel workflow implementations. Attendees will gain an overview of the latest BioBB features and learn how these advances facilitate the design, execution, and sharing of reproducible biomolecular simulation workflows across diverse computational environments.

Presenter

Adam Hospital

Adam Hospital is a Research Associate at the Molecular Modeling and Bioinformatics Unit (MMB) – Institute for Research in Biomedicine in Barcelona (IRB Barcelona) and Research Software Engineer for the Spanish National Institute of Bioinformatics (INB, ELIXIR-ES). From a computer science background, he jumped to the bioinformatics world and got trapped by the fascinating field of structural bioinformatics. He has been involved in scientific projects at IRB and at the Barcelona Supercomputing Center (BSC), and he has been leading the workflows team in the BioExcel CoE since its establishment in 2016. While working as a bioinformatician, he obtained his PhD in Biotechnology from the University of Barcelona with his thesis “High Throughput Computational Studies of Macromolecular Structure Flexibility”. He has developed and coordinated a set of public web servers and databases related to macromolecular structure flexibility, the last one being the Molecular Dynamics Data Bank (MDDB, MDposit).

X: @IRBBarcelona, @orozco_lab, @mddbEU

Bluesky: bsky.app/profile/irbbarcelona.org

https://www.linkedin.com/company/irbbarcelona/

https://www.facebook.com/irbbarcelona

https://www.instagram.com/irbbarcelona/

https://www.youtube.com/user/IRBBarcelona

View Slides on Zenodo

Abstract

Many biological processes are regulated by changes in protein covalent chemistry, from protonation and disulfide formation to post-translational modification. In this webinar, we will present our past and present work extending alchemical free-energy calculations in GROMACS/pmx to quantify the thermodynamic consequences of these changes. After a brief introduction to free-energy methods, we will discuss our ongoing work toward a general framework for covalent protein modifications and highlight some applications to biological questions including pH-dependent gating of aquaporins and modification-driven changes in amyloid fibril stability.

Presenter

Carter J. Wilson

Carter J. Wilson is a postdoctoral researcher in the Computational Biomolecular Dynamics group at the MPI-NAT in Göttingen, where he also completed his PhD. His work focuses on developing free-energy methods, with applications to membrane proteins and amyloid fibrils.

X: @wilsonjcarter

GitHub: wilsonjcarter

This special edition of the BioExcel webinar series features student speakers who were awarded poster prizes at the BioExcel Summer School 2026. Below you can find out more about our speakers and their research.

Marc El Hasbany

Marc is a first-year PhD student at the Institute of Chemistry of Nice (ICN), Université Côte d’Azur in France. He holds a Bachelor of Science in Bioinformatics from the Lebanese American University in Beirut, Lebanon, and completed a Master’s degree in In Silico Drug Design and Macromolecular Modelling at Université Paris Cité in Paris, where he was first introduced to molecular dynamics simulations. Under the supervision of Dr. Jérémie Topin, his doctoral research focuses on deciphering the molecular mechanisms of olfactory receptor activation and inhibition using molecular dynamics simulations, as well as identifying potential non-agonist ligands of ectopically expressed olfactory receptors. His work is conducted in collaboration with Seoul National University in South Korea, where computational predictions are experimentally validated

 Title:  Deciphering the molecular basis for activation and inhibition of olfactory receptors

Olfactory receptors (ORs) are part of the largest family of G protein-coupled receptors (GPCR) and constitute the molecular basis for the detection and discrimination of a vast chemical space of volatile compounds. ORs function through a combinatorial mechanism, meaning that a single odorant can activate multiple receptors, while a single receptor can respond to structurally diverse ligands. Beyond their sensory role, select ORs are ectopically expressed in non-olfactory tissues; Olfr412, for instance, is expressed in spermatozoa and participates in chemotaxis, highlighting the broader physiological relevance of this receptor family. Despite significant advances in the field regarding the molecular mechanisms of odorant recognition and receptor activation, receptor inhibition remains poorly understood.

This study integrates computational modeling with experimental validation to decipher and further understand the molecular determinants of OR activation and inhibition. Olfr412, the ectopically expressed OR in mice, bound with different types of odorants served as the selected model systems. We combine molecular docking and simulations to characterize agonist and non agonist binding at the atomic level. Comparative analysis of OR-agonist and OR-antagonist complexes reveals differences in receptor-ligand interactions highlighting the differences between activation and inhibition. Additionally, virtual screening was performed on a large dataset of odorants collected from the M2OR database to identify other potential agonists and non agonists. The selected odorants were subsequently tested experimentally to validate their effect on OR activity. Furthermore, Umbrella Sampling was employed to map the translocation pathways of odorants through the Olfr412.

Collectively, this work deepens our understanding of OR activation and inhibition at the molecular level, with implications for both their sensory and ectopic function, enabling the design of modulators.

José Luis Muñoz Reyes

José trained as a Chemical and Bioprocess Engineer at the Karlsruhe Institute of Technology (KIT), where he completed his Bachelor’s and Master’s degrees. After an internship and Master’s thesis at Boehringer Ingelheim, focusing on Raman spectroscopic quantification of protein mixtures, he joined the lab of Dr. Thales Kronenberger at the University of Tübingen for a PhD in computational drug discovery. His research focuses on molecular dynamics simulations, particularly enhanced-sampling methods, to understand the relationship between protein dynamics and function, and uncover allosteric mechanisms across protein families, including GPCRs, kinases, and membrane transporters.

Title: Understanding the activation dynamics of the CXCR4 receptor using adaptive sampling simulations

CXCR4 is a class A G protein-coupled receptor (GPCR) that plays a pivotal role in regulating the migration, positioning, and survival of diverse cell types. Aberrant CXCR4 signalling is implicated in numerous pathological conditions, including cancer, inflammation, autoimmune disorders, and HIV infection. Despite the availability of many experimental structures, the molecular determinants governing CXCR4 conformational transitions remain poorly understood, limiting the development of novel therapeutic strategies such as allosteric modulation. To characterize the dynamic landscape of CXCR4 and identify potential allosteric regulatory sites, we performed extensive molecular dynamics simulations spanning transitions between active and inactive receptor states. Our simulations reveal state-specific, highly persistent lipid insertion sites that represent prospective druggable pockets on the receptor surface. While conventional microsecond-scale simulations remained trapped in local free-energy minima, an adaptive sampling workflow based on the FAST algorithm efficiently promoted transitions between conformational states and enhanced exploration of the receptor’s conformational landscape. Using Markov state models, we show that CXCR4 activation proceeds through four metastable intermediates, providing a mechanistic framework for understanding receptor activation and identifying opportunities for allosteric drug design.

Gianluca Santini

Gianluca is a first-year Ph.D. student in Chemistry at the University of Milano-Bicocca, working within the molecular modeling research group. His doctoral project focuses on advancing Computer-Aided Drug Design by integrating Machine Learning (ML) with physics-based methods, particularly Molecular Dynamics (MD) simulations. The primary targets of his research are Nuclear Receptors, a crucial protein family involved in different pathophysiological conditions. By applying deep learning architectures to analyze MD trajectories, his work aims to decode the highly dynamic behavior of these receptors, isolating structural and functional signatures to guide the rational discovery of novel therapeutic modulators.

Title: Isolating Functional Signatures in Protein Dynamics with an Explainable Spatiotemporal Graph Neural Network

Biomolecular function is fundamentally driven by conformational dynamics, with proteins constantly adapting their structures in response to binding events or mutations. To understand these mechanisms, computational chemists frequently compare molecular dynamics (MD) simulations of proteins in distinct states (such as ligand-bound and unbound). The primary challenge in this comparative MD analysis lies in separating the true functional signatures from the stochastic thermal noise and simulation-specific variance. Standard geometric or dimensionality reduction methods often struggle to isolate these subtle, mechanistically relevant differences across independent simulation replicas.

In this talk, I will introduce an explainable deep learning approach (GISTnet-MD) designed to overcome these limitations and learn state-specific dynamic signatures directly from raw MD trajectories [1]. Specifically, I will illustrate how GISTnet-MD filters out trajectory-specific correlations and random fluctuations to capture the invariant biophysical patterns defining a specific functional state. By translating raw trajectory differences into interpretable, residue-level determinants, I will demonstrate how the model can successfully isolate fundamental mechanisms such as distributed allostery and specific activation signatures [1].

Finally, I will discuss the application of this framework to the nuclear receptor superfamily. To this end, I am constructing a comprehensive MD simulation dataset to systematically compare their active (agonist-bound) and inactive (apo) states, leveraging this AI-driven methodology to decode their allosteric regulation.

1. Motta, S. et al. Spatiotemporal Graph Neural Networks Reveal Conformational Binding Signature in Protein Dynamics. bioRxiv Preprint at https://doi.org/10.64898/2026.05.19.726195 (2026).

View slides on Zenodo

Abstract

Intrinsically disordered proteins and regions (collectively IDRs) are pervasive across proteomes in all kingdoms of life, help shape biological functions, and are involved in numerous diseases [1]. IDRs populate a diverse set of transiently formed structures yet defy commonly held sequence-structure-function relationships [1–3]. Recent developments in protein structure prediction have led to the ability to predict the three-dimensional structures of folded proteins at the proteome scale and have enabled large-scale studies of structure-function relationships. In contrast, knowledge of the conformational properties of fully disordered proteins and long disordered linkers is scarce, in part because the sequences of disordered proteins are poorly conserved and because only few have been characterized experimentally. In my talk I will describe how we can use molecular simulations with coarse-grained models and machine learning to study the relationship between sequence, conformational properties, and functions of IDRs [3].

First, I will describe how we have used experimental data on more than 100 proteins to learn a coarse-grained molecular energy function to predict conformational properties of IDRs [4,5]. By globally optimizing a transferable model, called CALVADOS, we can study the conformational ensemble of an IDR [5,6], multidomain protein [7] and IDRs interacting with disordered RNA [8]. I will describe the Bayesian formalism we developed to parameterize CALVADOS by targeting experimental data, and how this model enables us to study interactions within and between IDRs in biomolecular condensates [6–9].

Second, I will describe how CALVADOS makes it possible to perform large-scale simulations to explore the relationship between sequence, structure, and function of IDRs. I will describe how we have generated conformational ensembles of all intrinsically disordered regions of the human proteome and used these to provide insight into sequence-ensemble relationships and evolutionary conservation of IDR properties [10].

Third, I will describe how we can integrate AlphaFold with CALVADOS to generate conformational ensembles of proteins with both folded and disordered regions, how we benchmarked the method using data for hundreds of proteins, and how the speed and scalability enables us to construct conformational ensembles at the proteome level [11].

References

1. Holehouse, A. S., & Kragelund, B. B. (2023). The molecular basis for cellular function of intrinsically disordered protein regions. Nature Reviews Molecular Cell Biology, 1-25.
2. Lindorff-Larsen, K., & Kragelund, B. B. (2021). On the potential of machine learning to examine the relationship between sequence, structure, dynamics and function of intrinsically disordered proteins. Journal of Molecular Biology, 433(20), 167196.
3. von Bülow, S., Tesei, G., & Lindorff-Larsen, K. (2025). Machine learning methods to study sequence–ensemble–function relationships in disordered proteins. Current Opinion in Structural Biology, 92, 103028.
4. Norgaard, A. B., Ferkinghoff-Borg, J., & Lindorff-Larsen, K. (2008). Experimental parameterization of an energy function for the simulation of unfolded proteins. Biophysical Journal, 94(1), 182-192.
5. Tesei, G., Schulze, T. K., Crehuet, R., & Lindorff-Larsen, K. (2021). Accurate model of liquid–liquid phase behavior of intrinsically disordered proteins from optimization of single-chain properties. Proceedings of the National Academy of Sciences, 118(44), e2111696118.
6. Tesei, G., & Lindorff-Larsen, K. (2023). Improved predictions of phase behaviour of intrinsically disordered proteins by tuning the interaction range. Open Research Europe, 2, 94.
7. Cao, F., von Bülow, S., Tesei, G., & Lindorff‐Larsen, K. (2024). A coarse‐grained model for disordered and multi‐domain proteins. Protein Science, 33, e5172.
8. Yasuda, I., von Bülow, S., Tesei, G., Yamamoto, E., Yasuoka, K., & Lindorff-Larsen, K. (2025). Coarse-grained model of disordered RNA for simulations of biomolecular condensates. Journal of chemical theory and computation, 21(5), 2766-2779.
9. von Bülow, S., Tesei, G., Zaidi, F. K., Mittag, T., & Lindorff-Larsen, K. (2025). Prediction of phase-separation propensities of disordered proteins from sequence. Proceedings of the National Academy of Sciences, 122(13), e2417920122.
10. Tesei, G., Trolle, A. I., Jonsson, N., Betz, J., Knudsen, F.E., Pesce, F., Johansson, K. E., & Lindorff-Larsen, K. (2024). Conformational ensembles of the human intrinsically disordered proteome. Nature, 626(8000), 897-904.
11. von Bülow, S., Johansson, K. E., & Lindorff-Larsen, K. (2025). AF-CALVADOS: AlphaFold-guided simulations of multi-domain proteins at the proteome level. bioRxiv, 2025-10.

Presenters

Kresten Lindorff-Larsen

Kresten Lindorff-Larsen trained as a biochemist at the University of Copenhagen and Carlsberg Laboratory, and completed his Ph.D. at the University of Cambridge in 2004 under the supervision of Prof. Christopher M. Dobson. He then moved on to become an assistant professor in Copenhagen before joining D. E. Shaw Research in New York in 2007. He returned to Copenhagen in 2011, where he now serves as a Professor of Computational Protein Biophysics at the Linderstrøm-Lang Centre for Protein Science and is the director of the PRISM (Protein Interactions and Stability in Medicine and Genomics) centre.

Kresten received the Danish Independent Research Councils’ Young Researchers’ Award in 2006, was a co-recipient of the 2009 Gordon Bell Prize and the 2025 Novo Nordisk Foundation’s Prize for Natural Science Teachers at Universities, and has received several prestigious grants including a Hallas-Møller stipend (2011), a Sapere Aude starting grant (2012), a Novo Nordisk Foundation challenge programme grant (2019), and an ERC Synergy Grant (2023).

He was elected as a fellow of the Biophysical Society in 2024 and the same year he also became a member of the Royal Danish Academy of Sciences and Letters. Current research interests include developing and applying computational methods for studying the structure and dynamics of proteins, and the integration of biophysics and genomics research.

View slides on Zenodo

Abstract

Short peptides have been instrumental as models for more complex systems. Studying short peptides, we have gained understanding of fundamental folding events like alpha helix nucleation or hairpin folding, and in the determination of the microscopic origin of internal friction. Short peptides are also challenging tests for the calibration of modern force fields, and hence have been extensively used in recent optimization efforts. In my talk I will focus on our recent work on phase separation in our laboratory. Using peptide models, we test the hypothesis that phase separation is an emergent property determined by composition, even in the absence of a polypeptide chain, multivalency and patterning effects. For this, we use atomistic molecular dynamics (MD) simulations of saturated solutions of individual amino acids and mixtures thereof in stoichiometries comparable to those of phase-separating low com-plexity domains. Additionally, we disentangle the hierarchy of interaction strength between the two most dominant types of aromatic and positively charged residues in condensates. Our results are broadly consistent with trends observed in experiments and in atomistic simulations of full-length IDPs and reconcile findings from decades of work in physical chemistry and protein biophysics.

Presenters

David de Sancho

David is a research professor at the University of the Basque Country (UPV/EHU) and the Donostia International Physics Center, where he received tenure in 2023 following a Ramón y Cajal fellowship. His work is rooted in computational protein biophysics, a field he’s explored for 20 years by using molecular simulations to map out protein energy landscapes and modulate their functions. His journey started with a PhD focused on genetic algorithms and coarse-grained models, followed by postdoctoral stays at the Spanish Research Council and the University of Cambridge. Some of his key research achievements in this period were the determination of alpha-helix folding rates, resolving the molecular cause of internal friction and identifying gas tunnels in enzymes. After returning to Spain via Ikerbasque and CIC nanoGUNE, he now co-leads the BIOKT group, where they dive into the complex world of intrinsically disordered proteins using a mix of computational chemistry and molecular dynamics.

LinkedIn: @david-de-sancho-71b06150

Bluesky: @daviddesancho.bsky.social

View slides on Zenodo

Abstract

Biological membranes exhibit intricate, large-scale dynamical architectures that are intimately coupled to cellular function, and abnormal membrane organization is implicated in many diseases. These structures arise from collective molecular interactions. Although these interactions can be described using molecular dynamics simulations, the spatiotemporal scales required to capture membrane organization at cellular and subcellular scale remain beyond the practical reach of such approaches. Meanwhile, rapid advances in experimental techniques now provide unprecedented structural and dynamical detail. Three-dimensional electron microscopy enables visualization of membrane ultrastructure at near-molecular resolution, while super-resolution microscopy reveals membrane dynamics and organization in living cells. These developments create a growing need for computational frameworks capable of bridging molecular detail with mesoscale and cellular-scale phenomena. To address this challenge, I propose integrating mesoscale membrane models into a multiscale simulation framework for biomembranes through systematic mapping and backmapping schemes, enabling the transfer of structural and dynamical information across scales.

In this talk, I will first present a few representative cases demonstrating how large-scale membrane morphology and dynamics regulate molecular-scale processes. I will then introduce a mesoscale simulation framework implemented in the FreeDTS software package and describe its integration into a multiscale modeling pipeline through the TS2CG backmapping scheme. Finally, I will present our recent advances in data-driven mesoscale membrane modeling using the Helfrich Monte Carlo Flexible Fitting (HMFF) approach. This method enables the biasing of membrane simulations using high-resolution structural data from three-dimensional electron microscopy, allowing the generation of statistically meaningful ensembles of membrane morphologies. Together, FreeDTS, HMFF, TS2CG, and conventional molecular dynamics engines constitute a unified multiscale framework that enables simulations of biomembranes across scales.

Presenters

Weria Pezeshkian

Weria Pezeshkian has a master’s degree in condensed matter physics and a PhD in molecular biophysics. Currently, he is a group leader at the Niels Bohr Institute, University of Copenhagen. His group develops and applies multiscale computer simulation methods to explore cellular forms.

LinkedIn: @weria-pezeshkian-22015a51

Bluesky: @weria-lab.bsky.social

View slides on Zenodo

Abstract

Static structural models often fail to capture the dynamic mechanisms of protein interactions. To address this, we introduce DynaPIN, an open-source pipeline for extracting dynamic interface fingerprints from molecular simulations. DynaPIN unifies quality control metrics, interface prediction accuracy assessment, and atomistic interaction analysis into a single automated workflow, accessible at https://github.com/CSB-KaracaLab/DynaPIN. A key feature is our interface-specific analysis centered on a Dynamic Interface definition, which classifies residues based on the persistence of their interaction status over the simulation. We applied DynaPIN to representative rigid, medium, and difficult targets from the DynaBench dataset, an MD simulation resource for Docking Benchmark 5.5. Our results show that interface flexibility diverges from static accuracy classifications established in Docking Benchmark 5.5, as explored before. All in all, by providing standardized, frame-resolved outputs, DynaPIN’s aim is to facilitate mechanistic studies and generate standardized unbiased data for future dynamics-aware artificial intelligence models.

Presenters

Ezgi Karaca

Ezgi Karaca is a computational structural biologist specializing in structural modeling and biomolecular dynamics. She completed her Ph.D. (2013) at Utrecht University under the supervision of Prof. Alexandre Bonvin, where she advanced HADDOCK. She then conducted her postdoctoral research at EMBL Heidelberg (2013–2016) with Prof. Teresa Carlomagno and Prof. Orsolya Barabas, developing M3, a protein complex modeling tool capable of handling 20 molecules using sparse experimental data. In 2017, Dr. Karaca established the Computational Structural Biology Lab at Izmir Biomedicine and Genome Center and joined Dokuz Eylul University. Her lab develops classical and AI-driven approaches to model biomolecular interactions and their dynamics. She received an EMBO Installation Grant (2020). Dr. Karaca was also the first assessor from Türkiye in CASP14 and CASP15 for the assembly prediction rounds. Her lab’s first online tool, PROT-ON, a structure-based web server for designing interfacial mutations, gained significant attention from the scientific community, with 50,000 uses per year worldwide.

X: @Ezgi_Karaca_

Bluesky: @ezgikaraca.bsky.social

LinkedIn: @ezgi-karaca-ibg

Ayşe Berçin Barlas

Ayşe Berçin Barlas is a computational structural biologist specialized in molecular dynamics simulations and biomolecular interface analysis. She completed her Ph.D. at the Izmir Biomedicine and Genome Center (IBG) and Dokuz Eylül University under the supervision of Dr. Ezgi Karaca, where she investigated how biomolecular complexes encode functional specificity through dynamics. During her doctoral research, she developed a Comparative Dynamics Analysis (CDA) framework to dissect molecular basis of sequence-specific DNA recognition by methyltransferases. As a postdoctoral researcher at IBG, Dr. Barlas is currently working on the development of DynaPIN, a Python-based package for time-resolved characterization of protein-protein interfaces. She is also one of the contributors of DynaBench, a large-scale molecular dynamics dataset designed to capture interface dynamics beyond static reference structures.

View slides on Zenodo

Abstract

The study of the interior of biomolecules is still a challenging task due to the ephemeral character of tunnels, voids and cavities While experimental techniques provide invaluable structural information, they often offer limited access to the dynamic accessibility of internal pathways and molecular transport processes. Therefore, during the workshop we will be presenting a concise overview of several studies conducted over the past few years using solvent tracking software. This will be accompanied by an explanation of the technical background of the tool and practical training. The webinar will provide insight and understanding into the function and evolution of proteins, facilitating enzyme redesign and drug discovery. We will demonstrate how to access to information about protein interior and small molecules flow direction, how to identify regions in enzymes where small molecules are stuck or trapped. We will provide an examples, how the analysis of water behaviour can help with understand the enzymatic reaction mechanism, and how it can help with drug design and discovery.

Presenters

Artur Góra

Artur is a professor at Silesian University of Technology in Gliwce, Poland and head of the Tunneling Group, where AQUA-DUCT a software for small molecules tracking was developed.

Weronika Bagrowska

Weronika is a PhD candidate at the Silesian University of Technology and a member of the Tunneling Group. She has been involved in testing the AQUA-DUCT software and is currently focusing on exploring its potential applications in drug design projects.

Maria Bzówka

Maria obtained her PhD from the Silesian University of Technology, where she studied molecular aspects of protein regulation with a focus on the role of water molecules as mediators of intermolecular interactions. Currently, she is a postdoctoral researcher in the group of Francesco Luigi Gervasio (University of Geneva).

Group website: http://www.tunnelinggroup.pl/

AQUA-DUCT website: https://www.aquaduct.pl/