Source author record

Jérôme Hénin

Jérôme Hénin appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Biological Physics cond-mat.soft cond-mat.stat-mech Machine Learning physics.chem-ph physics.comp-ph

Catalog footprint

What is connected

3works

6topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

Molecular signatures of pressure-induced phase transitions in a lipid bilayer

Understanding how lipid bilayers respond to pressure is essential for interpreting the coupling between membrane proteins and their native environments. Here, we use all-atom molecular dynamics to examine the pressure-temperature behavior of model membranes composed of DMPC or $Δ$9-cis-PC. Within the studied range (288-308 K, 1-2000 bar), DMPC undergoes a liquid--gel transition, while $Δ$9-cis-PC remains fluid due to unsaturation. The CHARMM36 force field reproduces experimental boundaries with high fidelity: simulated DMPC transitions deviate by only 5-10 K and 100-300 bar, and $Δ$9-cis-PC exhibits no transition. Hysteresis is modest but most pronounced when starting from low-temperature gels. We identify area per lipid, bilayer thickness, and acyl-chain gauche fraction as sensitive phase markers; among these, the gauche fraction provides the most robust signature. Simulations indicate an interdigitated gel is the equilibrium structure under finite-size conditions. However, at low temperature and high pressure, interdigitation decreases, consistent with the experimental lamellar gel phase. This long-lived interdigitation critically impacts standard order parameters, specifically area per lipid and membrane thickness. These results underscore the accuracy of modern force fields and highlight how simulations mechanistically complement experimental studies of pressure-regulated membranes.

preprint2023arXiv

Designing losses for data-free training of normalizing flows on Boltzmann distributions

Generating a Boltzmann distribution in high dimension has recently been achieved with Normalizing Flows, which enable fast and exact computation of the generated density, and thus unbiased estimation of expectations. However, current implementations rely on accurate training data, which typically comes from computationally expensive simulations. There is therefore a clear incentive to train models with incomplete or no data by relying solely on the target density, which can be obtained from a physical energy model (up to a constant factor). For that purpose, we analyze the properties of standard losses based on Kullback-Leibler divergences. We showcase their limitations, in particular a strong propensity for mode collapse during optimization on high-dimensional distributions. We then propose strategies to alleviate these issues, most importantly a new loss function well-grounded in theory and with suitable optimization properties. Using as a benchmark the generation of 3D molecular configurations, we show on several tasks that, for the first time, imperfect pre-trained models can be further optimized in the absence of training data.

preprint2021arXiv

Human learning for molecular simulations: the Collective Variables Dashboard in VMD

The Collective Variables Dashboard is a software tool for real-time, seamless exploration of molecular structures and trajectories in a customizable space of collective variables. The Dashboard arises from the integration of the Collective Variables Module with the visualization software VMD, augmented with a fully discoverable graphical interface offering interactive workflows for the design and analysis of collective variables. Typical use cases include a priori design of collective variables for enhanced sampling and free energy simulations as well as post-mortem analysis of any type of simulation or collection of structures in a collective variable space. A combination of those cases commonly occurs when preliminary simulations, biased or unbiased, reveal that an optimized set of collective variables is necessary to improve sampling in further simulations. Then the Dashboard provides an efficient way to intuitively explore the space of likely collective variables, validate them on existing data, and use the resulting collective variable definitions directly in further biased simulations using the Collective Variables Module. We illustrate the use of the Dashboard on two applications: discovering coordinates to describe ligand unbinding from a protein binding site, and designing volume-based variables to bias the hydration of a transmembrane pore.