The ICTP-East African Institute for Fundamental Research (EAIFR) wishes to invite you to our next ICTP-EAIFR webinar. This seminar will take place on 20th November, 2025 and will be broadcast live on ZOOM at 4:00 PM Kigali TIme. It will also be recorded and later posted on the ICTP-EAIFR YouTube channel, where one can find all previous  ICTP-EAIFR webinar recordings.

Below are the details:

Speaker: Dr. Ariane Nunes-Alves, Junior Group Leader at the Technical University of Berlin (Germany)

Title: Integrating Molecular Dynamics Simulations and Machine Learning to Investigate Protein-Ligand Binding Kinetics

When: November 20, 2025 at 4:00 pm (Kigali time).

Register in advance for this meeting by clicking here.

All are very welcome.

Abstract: 

[NiFe] hydrogenases can act as efficient catalysts for hydrogen oxidation and biofuel production. However, their industrial use is limited due to inhibition by gas molecules present in the environment, such as O₂ and CO. Here, we combined unbiased molecular dynamics (MD) simulations, enhanced sampling with tau-Random Accelerated molecular dynamics (tauRAMD) and machine learning (ML) to investigate binding and unbinding paths of small gas molecules (substrate, H₂, and inhibitors, O₂ and CO) in [NiFe] hydrogenases. Unbiased MD simulations captured multiple H₂ binding and unbinding events, reproducing experimental association rates and revealing symmetry between entry and exit pathways for H₂ (1). tauRAMD simulations provided relative residence times in agreement with experiments for the dissociation of CO from WT and 10 different mutants of [NiFe] hydrogenase (2). Data analysis of these simulations revealed a key bottleneck, formed by residues V74 and L122, whose conformational shifts modulate ligand access, and path probabilities for the different ligands. While the most probable pathways are the same, the secondary pathways are different for substrate and inhibitors.
Finally, we developed PathInHydro (3), a set of supervised ML models trained on CO and H₂ unbinding trajectories, which automates pathway identification in MD simulations and generalizes to additional ligands and hydrogenases. Together, these results provide mechanistic understanding of gas diffusion in [NiFe] hydrogenases and suggest strategies for engineering O₂- and CO-tolerant enzymes.

References
1. Sohraby, F.; Nunes-Alves, A. Symmetric ligand binding pathways and dual-state bottleneck in [NiFe] hydrogenases from unbiased molecular dynamics. J. Phys. Chem. Lett. 16: 7960–7967, 2025.
2. Sohraby, F.; Nunes-Alves, A. Characterization of the bottlenecks and pathways for inhibitor dissociation from [NiFe] hydrogenase. J. Chem. Inf. Model., 64: 4193-4203, 2024.
3. Sohraby, F.; Guo, J.-Y.; Nunes-Alves, A. PathInHydro, a set of machine learning models to identify unbinding pathways of gas molecules in [NiFe] hydrogenases. J. Chem. Inf. Model., 65:589-602, 2025.

Biography:

Dr. Ariane Nunes-Alves obtained her bachelor's degree in Biology at the University of Sao Paulo (Brazil), and a master's degree and a PhD in Biochemistry at the University of Sao Paulo (Brazil). From 2018 to 2021 she was a postdoctoral researcher at the Heidelberg Institute for Theoretical Studies (Germany), with funding from the Alexander von Humboldt foundation. Since 2021 Dr. Nunes-Alves has been a Junior Group Leader at the Technical University of Berlin (Germany). The focus of her research group is to develop and apply computational methods to investigate protein-ligand binding in vitro and in vivo. Dr. Nunes-Alves is an Associate Editor of the Journal of Chemical Information and Modeling (part of the American Chemical Society).