arXiv subjects

Luigi Bonati

Publications and source records attributed to Luigi Bonati.

At least 19 recordsLinked to original sources

arXiv2026

SelfTICA: contrastive learning of dynamical representations for rare-event sampling and characterization

Rare events govern many important molecular processes but remain difficult to characterize within accessible simulation timescales. This challenge has motivated machine-learning methods to construct low-dimensional collective variables for specific objectives, including state discrimination, slow-mode identification, and committor-function approximation. Here we introduce SelfTICA, a self-supervised framework that uses contrastive learning on time-lagged configurations to learn a latent representation of the slow modes governing relevant transitions. Once learned, this representation is frozen and reused across downstream tasks, including collective variables construction, enhanced sampling, free-energy estimation, and committor learning. Compared with direct slow-mode optimization, SelfTICA improves training stability and provides collective variables even from limited and exploratory trajectories, accelerating rare-event sampling and enabling accurate free-energy estimation. The dynamical information encoded during pretraining also accelerates committor learning and enables characterization of transition-state regions. These results show that contrastive learning of slow dynamical representations provides a common foundation for rare-event sampling and mechanistic analysis.

physics.comp-ph↗

arXiv2025

Electrochemical Interfaces at Constant Potential: Data-Efficient Transfer Learning for Machine-Learning-Based Molecular Dynamics

Simulating electrified metal/water interfaces with explicit solvent under constant potential is essential for understanding electrochemical processes, yet remains prohibitively expensive with ab initio methods. We present TRECI, a data-efficient workflow for constructing machine learning force-fields (ML-FFs) that achieve ab initio-level accuracy in electronically grand-canonical molecular dynamics. By leveraging transfer learning from general-purpose and domain-specific models, TRECI enables stable and accurate simulations across a wide potential range using a reduced number of reference configurations. This efficiency allows the use of high-level meta-GGA functionals and rigorous surface-electrification schemes. Applied to Cu(111)/water, models trained on just one thousand configurations yield accurate molecular dynamics simulations, capturing bias-dependent solvent restructuring effects not previously reported. TRECI offers a general strategy for characterising diverse materials and interfacial chemistries, significantly lowering the cost of realistic constant-potential simulations and expanding access to quantitative electrochemical modelling.

physics.comp-ph↗

arXiv2025

Enhanced Sampling in the Age of Machine Learning: Algorithms and Applications

Molecular dynamics simulations hold great promise for providing insight into the microscopic behavior of complex molecular systems. However, their effectiveness is often constrained by long timescales associated with rare events. Enhanced sampling methods have been developed to address these challenges, and recent years have seen a growing integration with machine learning techniques. This review provides a comprehensive overview of how they are reshaping the field, with a particular focus on the data-driven construction of collective variables. Furthermore, these techniques have also improved biasing schemes and unlocked novel strategies via reinforcement learning and generative approaches. In addition to methodological advances, we highlight applications spanning different areas such as biomolecular processes, ligand binding, catalytic reactions, and phase transitions. We conclude by outlining future directions aimed at enabling more automated strategies for rare-event sampling.

physics.comp-ph↗

arXiv2025

Self-Supervised Evolution Operator Learning for High-Dimensional Dynamical Systems

We introduce an encoder-only approach to learn the evolution operators of large-scale non-linear dynamical systems, such as those describing complex natural phenomena. Evolution operators are particularly well-suited for analyzing systems that exhibit complex spatio-temporal patterns and have become a key analytical tool across various scientific communities. As terabyte-scale weather datasets and simulation tools capable of running millions of molecular dynamics steps per day are becoming commodities, our approach provides an effective tool to make sense of them from a data-driven perspective. The core of it lies in a remarkable connection between self-supervised representation learning methods and the recently established learning theory of evolution operators. To show the usefulness of the proposed method, we test it across multiple scientific domains: explaining the folding dynamics of small proteins, the binding process of drug-like molecules in host sites, and autonomously finding patterns in climate data. Code and data to reproduce the experiments are made available open source.

cs.LG↗

arXiv2025

Weighted Active Space Protocol for Multireference Machine-Learned Potentials

Multireference methods such as multiconfiguration pair-density functional theory (MC-PDFT) offer an effective means of capturing electronic correlation in systems with significant multiconfigurational character. However, their application to train machine learning-based interatomic potentials (MLPs) for catalytic dynamics has been challenging due to the sensitivity of multireference calculations to the underlying active space, which complicates achieving consistent energies and gradients across diverse nuclear configurations. To overcome this limitation, we introduce the Weighted Active-Space Protocol (WASP), a systematic approach to assign a consistent active space for a given system across uncorrelated configurations. By integrating WASP with MLPs and enhanced sampling techniques, we propose a data-efficient active learning cycle that enables the training of an MLP on multireference data. We demonstrate the method on the TiC+-catalyzed C-H activation of methane, a reaction that poses challenges for Kohn-Sham density functional theory due to its significant multireference character. This framework enables accurate and efficient modeling of catalytic dynamics, establishing a new paradigm for simulating complex reactive processes beyond the limits of conventional electronic-structure methods.

physics.chem-ph↗

arXiv2025

Fast and Fourier Features for Transfer Learning of Interatomic Potentials

Training machine learning interatomic potentials that are both computationally and data-efficient is a key challenge for enabling their routine use in atomistic simulations. To this effect, we introduce franken, a scalable and lightweight transfer learning framework that extracts atomic descriptors from pretrained graph neural networks and transfer them to new systems using random Fourier features-an efficient and scalable approximation of kernel methods. Franken enables fast and accurate adaptation of general-purpose potentials to new systems or levels of quantum mechanical theory without requiring hyperparameter tuning or architectural modifications. On a benchmark dataset of 27 transition metals, franken outperforms optimized kernel-based methods in both training time and accuracy, reducing model training from tens of hours to minutes on a single GPU. We further demonstrate the framework's strong data-efficiency by training stable and accurate potentials for bulk water and the Pt(111)/water interface using just tens of training structures. Our open-source implementation (https://franken.readthedocs.io) offers a fast and practical solution for training potentials and deploying them for molecular dynamics simulations across diverse systems.

physics.comp-ph↗

arXiv2024

PLUMED Tutorials: a collaborative, community-driven learning ecosystem

In computational physics, chemistry, and biology, the implementation of new techniques in a shared and open source software lowers barriers to entry and promotes rapid scientific progress. However, effectively training new software users presents several challenges. Common methods like direct knowledge transfer and in-person workshops are limited in reach and comprehensiveness. Furthermore, while the COVID-19 pandemic highlighted the benefits of online training, traditional online tutorials can quickly become outdated and may not cover all the software's functionalities. To address these issues, here we introduce ``PLUMED Tutorials'', a collaborative model for developing, sharing, and updating online tutorials. This initiative utilizes repository management and continuous integration to ensure compatibility with software updates. Moreover, the tutorials are interconnected to form a structured learning path and are enriched with automatic annotations to provide broader context. This paper illustrates the development, features, and advantages of PLUMED Tutorials, aiming to foster an open community for creating and sharing educational resources.

physics.ed-ph↗

arXiv2024

Advanced simulations with PLUMED: OPES and Machine Learning Collective Variables

Many biological processes occur on time scales longer than those accessible to molecular dynamics simulations. Identifying collective variables (CVs) and introducing an external potential to accelerate them is a popular approach to address this problem. In particular, $\texttt{PLUMED}$ is a community-developed library that implements several methods for CV-based enhanced sampling. This chapter discusses two recent developments that have gained popularity in recent years. The first is the On-the-fly Probability Enhanced Sampling (OPES) method as a biasing scheme. This provides a unified approach to enhanced sampling able to cover many different scenarios: from free energy convergence to the discovery of metastable states, from rate calculation to generalized ensemble simulation. The second development concerns the use of machine learning (ML) approaches to determine CVs by learning the relevant variables directly from simulation data. The construction of these variables is facilitated by the $\texttt{mlcolvar}$ library, which allows them to be optimized in Python and then used to enhance sampling thanks to a native interface inside $\texttt{PLUMED}$. For each of these methods, in addition to a brief introduction, we provide guidelines, practical suggestions and point to examples from the literature to facilitate their use in the study of the process of interest.

physics.comp-ph↗

arXiv2024

Descriptors-free Collective Variables From Geometric Graph Neural Networks

Enhanced sampling simulations make the computational study of rare events feasible. A large family of such methods crucially depends on the definition of some collective variables (CVs) that could provide a low-dimensional representation of the relevant physics of the process. Recently, many methods have been proposed to semi-automatize the CV design by using machine learning tools to learn the variables directly from the simulation data. However, most methods are based on feed-forward neural networks and require as input some user-defined physical descriptors. Here, we propose to bypass this step using a graph neural network to directly use the atomic coordinates as input for the CV model. This way, we achieve a fully automatic approach to CV determination that provides variables invariant under the relevant symmetries, especially the permutational one. Furthermore, we provide different analysis tools to favor the physical interpretation of the final CV. We prove the robustness of our approach using different methods from the literature for the optimization of the CV, and we prove its efficacy on several systems, including a small peptide, an ion dissociation in explicit solvent, and a simple chemical reaction.

physics.comp-ph↗

arXiv2024

Combining transition path sampling with data-driven collective variables through a reactivity-biased shooting algorithm

Rare event sampling is a central problem in modern computational chemistry research. Among the existing methods, transition path sampling (TPS) can generate unbiased representations of reaction processes. However, its efficiency depends on the ability to generate reactive trial paths, which in turn depends on the quality of the shooting algorithm used. We propose a new algorithm based on the shooting success rate, i.e. reactivity, measured as a function of a reduced set of collective variables (CVs). These variables are extracted with a machine learning approach directly from TPS simulations, using a multi-task objective function. Iteratively, this workflow significantly improves shooting efficiency without any prior knowledge of the process. In addition, the optimized CVs can be used with biased enhanced sampling methodologies to accurately reconstruct the free energy profiles. We tested the method on three different systems: a two-dimensional toy model, conformational transitions of alanine dipeptide, and hydrolysis of acetyl chloride in bulk water. In the latter, we integrated our workflow with an active learning scheme to learn a reactive machine learning-based potential, which allowed us to study the mechanism and free energy profile with an ab initio-like accuracy.

physics.comp-ph↗

arXiv2024

Unraveling the Crystallization Kinetics of the Ge$_2$Sb$_2$Te$_5$ Phase Change Compound with a Machine-Learned Interatomic Potential

The phase change compound Ge$_2$Sb$_2$Te$_5$ (GST225) is exploited in advanced non-volatile electronic memories and in neuromorphic devices which both rely on a fast and reversible transition between the crystalline and amorphous phases induced by Joule heating. The crystallization kinetics of GST225 is a key functional feature for the operation of these devices. We report here on the development of a machine-learned interatomic potential for GST225 that allowed us to perform large scale molecular dynamics simulations (over 10000 atoms for over 100 ns) to uncover the details of the crystallization kinetics in a wide range of temperatures of interest for the programming of the devices. The potential is obtained by fitting with a deep neural network (NN) scheme a large quantum-mechanical database generated within Density Functional Theory. The availability of a highly efficient and yet highly accurate NN potential opens the possibility to simulate phase change materials at the length and time scales of the real devices.

cond-mat.mtrl-sci↗

arXiv2024

Deep learning path-like collective variable for enhanced sampling molecular dynamics

Several enhanced sampling techniques rely on the definition of collective variables to effectively explore free energy landscapes. Existing variables that describe the progression along a reactive pathway offer an elegant solution but face a number of limitations. In this paper, we address these challenges by introducing a new path-like collective variable called the `Deep-locally-non-linear-embedding', which is inspired by principles of the locally linear embedding technique and is trained on a reactive trajectory. The variable mimics the ideal reaction coordinate by automatically generating a non-linear combination of features through a differentiable generalized autoencoder that combines a neural network with a continuous k-nearest-neighbor selection. Among the key advantages of this method is its capability to automatically choose the metric for searching neighbors and to learn the path from state A to state B without the need to handpick landmarks a priori. We demonstrate the effectiveness of DeepLNE by showing that the progression along the path variable closely approximates the ideal reaction coordinate in toy models such as the Müller-Brown-potential and alanine dipeptide. We then use it in molecular dynamics simulations of an RNA tetraloop, where we highlight its capability to accelerate transitions and converge the free energy of folding.

physics.chem-ph↗

arXiv2024

Transfer learning for atomistic simulations using GNNs and kernel mean embeddings

Interatomic potentials learned using machine learning methods have been successfully applied to atomistic simulations. However, accurate models require large training datasets, while generating reference calculations is computationally demanding. To bypass this difficulty, we propose a transfer learning algorithm that leverages the ability of graph neural networks (GNNs) to represent chemical environments together with kernel mean embeddings. We extract a feature map from GNNs pre-trained on the OC20 dataset and use it to learn the potential energy surface from system-specific datasets of catalytic processes. Our method is further enhanced by incorporating into the kernel the chemical species information, resulting in improved performance and interpretability. We test our approach on a series of realistic datasets of increasing complexity, showing excellent generalization and transferability performance, and improving on methods that rely on GNNs or ridge regression alone, as well as similar fine-tuning approaches.

cs.LG↗

arXiv2023

A unified framework for machine learning collective variables for enhanced sampling simulations: $\texttt{mlcolvar}$

Identifying a reduced set of collective variables is critical for understanding atomistic simulations and accelerating them through enhanced sampling techniques. Recently, several methods have been proposed to learn these variables directly from atomistic data. Depending on the type of data available, the learning process can be framed as dimensionality reduction, classification of metastable states or identification of slow modes. Here we present $\texttt{mlcolvar}$, a Python library that simplifies the construction of these variables and their use in the context of enhanced sampling through a contributed interface to the PLUMED software. The library is organized modularly to facilitate the extension and cross-contamination of these methodologies. In this spirit, we developed a general multi-task learning framework in which multiple objective functions and data from different simulations can be combined to improve the collective variables. The library's versatility is demonstrated through simple examples that are prototypical of realistic scenarios.

physics.comp-ph↗

arXiv2022

Characterizing metastable states with the help of machine learning

Present-day atomistic simulations generate long trajectories of ever more complex systems. Analyzing these data, discovering metastable states, and uncovering their nature is becoming increasingly challenging. In this paper, we first use the variational approach to conformation dynamics to discover the slowest dynamical modes of the simulations. This allows the different metastable states of the system to be located and organized hierarchically. The physical descriptors that characterize metastable states are discovered by means of a machine learning method. We show in the cases of two proteins, Chignolin and Bovine Pancreatic Trypsin Inhibitor, how such analysis can be effortlessly performed in a matter of seconds. Another strength of our approach is that it can be applied to the analysis of both unbiased and biased simulations.

physics.comp-ph↗

arXiv2021

Deep learning the slow modes for rare events sampling

The development of enhanced sampling methods has greatly extended the scope of atomistic simulations, allowing long-time phenomena to be studied with accessible computational resources. Many such methods rely on the identification of an appropriate set of collective variables. These are meant to describe the system's modes that most slowly approach equilibrium. Once identified, the equilibration of these modes is accelerated by the enhanced sampling method of choice. An attractive way of determining the collective variables is to relate them to the eigenfunctions and eigenvalues of the transfer operator. Unfortunately, this requires knowing the long-term dynamics of the system beforehand, which is generally not available. However, we have recently shown that it is indeed possible to determine efficient collective variables starting from biased simulations. In this paper, we bring the power of machine learning and the efficiency of the recently developed on-the-fly probability enhanced sampling method to bear on this approach. The result is a powerful and robust algorithm that, given an initial enhanced sampling simulation performed with trial collective variables or generalized ensembles, extracts transfer operator eigenfunctions using a neural network ansatz and then accelerates them to promote sampling of rare events. To illustrate the generality of this approach we apply it to several systems, ranging from the conformational transition of a small molecule to the folding of a mini-protein and the study of materials crystallization.

physics.comp-ph↗

arXiv2021

Using metadynamics to build neural network potentials for reactive events: the case of urea decomposition in water

The study of chemical reactions in aqueous media is very important for its implications in several fields of science, from biology to industrial processes. Modelling these reactions is however difficult when water directly participates in the reaction. Since it requires a fully quantum mechanical description of the system, $\textit{ab-initio}$ molecular dynamics is the ideal candidate to shed light on these processes. However, its scope is limited by a high computational cost. A popular alternative is to perform molecular dynamics simulations powered by machine learning potentials, trained on an extensive set of quantum mechanical calculations. Doing so reliably for reactive processes is difficult because it requires including very many intermediate and transition state configurations. In this study, we used an active learning procedure accelerated by enhanced sampling to harvest such structures and to build a neural-network potential to study the urea decomposition process in water. This allowed us to obtain the free energy profiles of this important reaction in a wide range of temperatures, to discover a number of novel metastable states and to improve the accuracy of the kinetic rates calculations. Furthermore, we found that the formation of the zwitterionic intermediate has the same probability of occurring via an acidic or a basic pathway, which could be the cause of the insensitivity of reaction rates to the pH solution.

physics.chem-ph↗

arXiv2021

Training collective variables for enhanced sampling via neural networks based discriminant analysis

A popular way to accelerate the sampling of rare events in molecular dynamics simulations is to introduce a potential that increases the fluctuations of selected collective variables. For this strategy to be successful, it is critical to choose appropriate variables. Here we review some recent developments in the data-driven design of collective variables, with a focus on the combination of Fisher's discriminant analysis and neural networks. This approach allows to compress the fluctuations of metastable states into a low-dimensional representation. We illustrate through several examples the effectiveness of this method in accelerating the sampling, while also identifying the physical descriptors that undergo the most significant changes in the process.

physics.comp-ph↗