Searcharxiv⌕ Search

arXiv subjects

Rafael Gómez-Bombarelli

Publications and source records attributed to Rafael Gómez-Bombarelli.

At least 19 recordsLinked to original sources
arXiv2026

Zatom-2: Multitask Pretraining on Atomistic Data for Generative Modeling across Domains

Unified atomistic modeling has the potential to accelerate discovery in chemistry, materials science, and biology by bridging data-rich chemical domains and data-scarce biological contexts. However, existing generative approaches to atomistic modeling remain highly specialized to scientific disciplines (chemistry vs. biology) or do not leverage both high-volume organic (molecule) and inorganic (material) data for general-purpose pretraining. To this end, we introduce Zatom-2, an atomistic generative model pretrained on approximately five million structures from the OMol25 and OMat24 electronic structure datasets. Zatom-2 features a multiscale Transformer architecture coupled with conditional flow matching that supports force conditioning and foundational pretraining tasks such as generation, structure prediction, and prediction of molecular and material energies and forces. Empirically, Zatom-2 achieves better molecular distribution fidelity than Zatom-1 and achieves strong performance on existing molecule and material generation benchmarks. Zatom-2 demonstrates the ability to control sample generation across low- and high-force regimes, and enhances protein generation in a low-data setting through joint generative-predictive pretraining and transfer learning, increasing protein backbone designability in a length extrapolation setting from 67.8% without pretraining to 74.8% after finetuning on 2,000 protein domains.

cs.LG↗
arXiv2026

Learning Lattice Parameters from Powder X-Ray Diffraction Data Using Invariants

We present a machine learning (ML) method to determine unit cell parameters from powder X-Ray diffraction (XRD) data using a novel invariant lattice representation. In ML, the data representation used can have a substantial impact on the prediction quality. Previous approaches have directly predicted lattice parameters ($a,b,c,α,β,γ$) from XRD inputs. However, these parameters depend strongly on the unit cell reduction or convention used. In this work, we construct an invariant representation of the reciprocal lattice that is independent of primitive cell convention, based on the bispectrum--a descriptor built from spherical harmonic projections of lattice points. The calculation of the lattice bispectrum is differentiable, and we demonstrate how to invert it using gradient-based optimization initialized from a nearest-neighbor lookup. We show that with a fixed ML model architecture, using the lattice bispectrum as the ML target rather than the unit cell parameters leads to more accurate lattice parameter predictions. For example, using the MP-20 dataset, the bispectrum reduces length mean absolute percentage error (MAPE) from 11.08% to 2.43% and angle MAPE from 12.15% to 2.45% compared to direct prediction with the same model architecture. We additionally benchmark our approach against pre-existing XRD to crystal structure models such as Crystalyze and assess its performance on experimental data. Beyond unit cell representation, we anticipate this invariant lattice representation could serve more broadly as a geometry-aware target for other crystallographic machine learning tasks such as structure generation.

physics.comp-ph↗
arXiv2026

LLM sequential decision making under uncertainty in biochemical domains

Large language models (LLMs) are increasingly used to drive scientific discovery. Understanding how LLMs make decisions from new data and memory of the literature is vital before trusting them to design experiments under tight experimental budgets. However, their decision strategies are invisible in the current performance scores used to evaluate research agents. Here, we benchmark five frontier LLMs in a Bayesian Optimization setting against published statistical baselines on seven combinatorial datasets spanning protein engineering, reaction optimization, molecular design, peptide self-assembly, and catalysis. Performance is paired with direct measurements of model beliefs and actions, enabling highly resolved behavior analysis. A prompt ablation that progressively strips context separates memorization from chemical reasoning and from bare categorical optimization. Prior chemical knowledge helps in expectation, but with high variance and occasionally even harms performance. No configuration tested decisively beats a mean statistical baseline across domains. Belief-movement and Martingale diagnostics, corrected here for a measurement-noise bias that mislabels rational agents as irrational, show that models overreact to incoming data rather than entrenching on their priors in the contexts studied here. Interestingly, while LLM actions are exploitative, models sincerely intend to explore and consistently act on that intent. This failure is a competence gap arising from context-stickiness. Removing in-context history restores exploration, indicating that priors and data must be decoupled to achieve effective LLM-driven discovery.

cs.AI↗
arXiv2026

AI-guided high-throughput discovery of iridium- and ruthenium-free palladium-oxide catalysts for durable acidic oxygen evolution

Catalyzing acidic oxygen evolution at the proton-exchange-membrane water electrolysis (PEMWE) anode relies almost entirely on iridium or ruthenium, drawn from concentrated supply chains that constrain gigawatt-scale deployment. We report an artificial intelligence (AI)-guided, human-supervised closed-loop platform (>90% automation) integrating combinatorial sputter synthesis, high-throughput screening, machine-learning composition-property models, adaptive multi-objective optimization, and context-aware large-language-model reasoning, where lead catalysts advanced to long-term validation in 1 M H2SO4 at 10 mA cm-2. Navigating a combinatorial metal oxide space, the platform iteratively evaluated the activity-stability trade-off of 2,942 catalysts across 53 material systems and 26 elements, surfacing Ir- and Ru-free complex oxides such as InMnPdOx and NiTaPdOx that conventional design logic, and off-the-shelf language models, would not predict. In retrospective benchmarking, our sequential learning agent advanced the activity-stability frontier faster than fixed-policy Bayesian optimization or in-context language-model selection. During long-term testing, NiTaPdOx operated at lower overpotential than PdOx, but both eventually exceeded 0.5 V: PdOx at ~200 h and NiTaPdOx at ~470 h. InMnPdOx showed a similar overpotential improvement in addition to a dramatic increase in operational stability, retaining overpotential below 0.5 V over 1,000 h of operation. The additive elements promote the formation of a nanostructure that is associated with catalytic activity while stabilizing Pd against corrosion. The results highlight the power of AI-driven science in addressing long-standing challenges in materials chemistry, and the greater availability of Pd relative to incumbent Ir and Ru offers a near-term option to ease supply constraints on scaled electrochemical H2 generation.

cond-mat.mtrl-sci↗
arXiv2026

Vibrational, structural, and chemical fingerprints of ion diffusion in crystalline solids

Predicting mobile-ion self-diffusivity $D^*$ from molecular dynamics (MD) simulations is essential for identifying promising solid-state electrolytes, but directly simulating ion diffusion is computationally expensive, particularly with high-accuracy machine learning interatomic potentials (MLIPs). Diffusion is a slow, emergent process that requires long trajectories to converge. Thermodynamic properties, by contrast, converge much faster: the enthalpy $h$, vibrational entropy $s_{vib}$, and 2-body, excess configurational entropy $s^{ex}_{2,config}$ can be extracted from comparatively short MD trajectories, and they encode rich information about the free energy landscape from which transport properties like self-diffusivity ultimately arise. Intuitive correlations are discussed between these thermodynamic properties and ion diffusion, motivating a data-driven approach to exploit this link. A simple neural network was trained to predict diffusivity from features computed over short MD trajectories: a vibrational fingerprint (the vibrational density of states, VDOS) and a structural fingerprint (the radial distribution function, RDF), conditioned on chemistry information encoded in the MLIP embedding. This combination allows the model to predict the converged $\log_{10} D^*$ (cm$^2$/s) \textemdash\ normally obtained from significantly longer MD simulations \textemdash\ with a mean absolute error of 0.398 and a Spearman's rank correlation $ρ$ of 0.844.

cond-mat.mtrl-sci↗
arXiv2026

Universal Thermodynamic Interatomic Potentials for Crystalline Materials

Free energies govern solid-state phase stability, yet computational materials discovery still relies largely on ground-state energies because free energy calculations require ensemble averages. We introduce the thermodynamic interatomic potential (TIP), which extends an interatomic potential from its static energy to a thermodynamically consistent Gibbs free energy model, with thermodynamic responses following from temperature and pressure by automatic differentiation. We implement TIP[UMA] using the universal potential UMA, train it on free energies from quasi-harmonic to molecular dynamics fidelity, and calibrate it to higher-resolution calculations or experiment. From a single evaluation, it returns the equation of state of a crystal and locates phase transitions among competing branches, including dynamically stabilized phases. Fine-tuning extends the model to alloy solubility limits and miscibility gaps. TIP makes the free energy as accessible as the potential energy, opening finite-temperature phase stability to high-throughput discovery.

cond-mat.mtrl-sci↗
arXiv2026

Microstructural Insights into Fast Ion Transport in Solid Electrolytes via Multiscale Modeling

Improving solid electrolytes is critical for high-performance all-solid-state batteries, yet the microstructural features that enable fast ion transport remain poorly understood. Here, we use multiscale modeling to resolve polycrystalline ion transport from atomic-scale hopping at grain boundaries to continuum-scale percolation, thereby providing insights into realistic solid-electrolyte microstructures. Accurate lightweight machine-learning potentials -- developed via closed-loop active learning for exemplar argyrodites Li$_6$PS$_5$X, X $\in$ {Cl, Br, I} -- are employed to integrate molecular dynamics with finite element simulations. We find that diffusion barriers of the anion-ordered bulk scale linearly with anion radius. Grain boundaries exert opposite effects depending on the bulk: enhancing ion diffusion in low-diffusivity phases but suppressing it in fast-diffusing ones. Li$_6$PS$_5$I exhibits non-Arrhenius transport behavior consistent with experimental observations. Our results clarify the pivotal role of grain boundaries in ion transport and guide a priori microstructural design of advanced solid electrolytes.

cond-mat.mtrl-sci↗
arXiv2026

Atomistic Language Models Understand and Generate Materials

Atomistic structure and natural language have long been modeled separately, with language models either calling atomistic models as tools or being fine-tuned on lossy textual encodings that discard atomistic information. We introduce Atomistic Language Models (ALMs) to pursue native multimodality, in which a single language backbone understands atomistic structures, generates materials from natural language, and optimizes crystal structures as instructed by text. By unifying a pretrained atomistic encoder, large language model, and denoising diffusion model through purely continuous projectors and staged training, ALMs achieve state-of-the-art results on crystal structure prediction and de novo generation. ALMs are enabled by a continuous bridge that maps language model embeddings directly into the steering space of atomistic diffusion, and are assisted by Text-to-Crystal Feynman-Kac (T2C-FK), a particle-based sampler that scores partial denoising trajectories to enforce stoichiometric targets at inference time. To evaluate the ability of ALMs to optimize and generate materials from natural-language prompts and 3D atom-coordinate inputs, we introduce ALM Bench, the first benchmark for text-conditioned crystal generation and optimization. Code, training data, and model weights will be released soon.

cs.LG↗
arXiv2026

Few-step Cofolding with All-Atom Flow Maps

All-atom generative modeling of 3D biomolecular complexes has emerged as the dominant paradigm for predicting the structure of proteins and protein-ligand systems. Generating structures at the atomic level of fidelity, however, typically requires expensive iterative diffusion rollouts, making both conventional deployment and inference-time search techniques computationally costly. In this paper, we introduce the Denoiser Cofolding All-Atom Flowmap (DeCAF) framework for distilling state-of-the-art all-atom cofolding models into all-atom flow maps that produce high-quality samples in only a few inference steps. We build DeCAF on a denoiser-based formulation of flow maps with endpoint losses that naturally support SE(3) rigid alignment, which we show is critical for training accurate models. We further derive a simple change of variables that lets DeCAF operate in the σ-space noise schedule of EDM-style architectures, enabling direct distillation from pretrained cofolding diffusion models. Equipped with DeCAF's flowmap lookahead, we introduce a purpose-built inference-time framework that improves sampling through reward-guided search. Empirically, DeCAF-Boltz statistically improves over Boltz-1x in both accuracy (RMSD) and physical validity scores of protein-ligand poses at strict NFE budgets on the challenging Runs N' Poses, while also showing a more optimal Pareto frontier across all inference compute budgets on PoseBusters. Distilling the state-of-the-art Pearl cofolding model, DeCAF-Pearl outperforms diffusion-based cofolding models and matches its teacher on success rate while using 5x fewer NFEs. We release our code at https://github.com/genesistherapeutics/decaf.

cs.LG↗
arXiv2026

XRDiff: Crystal Structure Prediction from Powder X-Ray Diffraction Data Using Diffusion Models

Determining the crystal structure of a material from its powder X-ray diffraction (PXRD) pattern is a central challenge in materials science. PXRD is an accessible and widely used characterization technique, yet recovering the atomic structure from diffraction data requires solving an underdetermined inverse problem due to the loss of phase information. Generative modeling can provide a prior over atomic structure and learn the mapping from PXRD patterns to crystal structures via simulated structure-spectrum pairs. We present XRDiff, a diffusion model that recovers crystal structures from PXRD given either the stoichiometry or, in a more challenging setting, the elemental constituents and total number of atoms in the unit cell. We evaluate on datasets where each stoichiometry has multiple polymorphs and all polymorphs of a given composition are held out together, ensuring that high performance reflects genuine use of the diffraction signal. XRDiff achieves strong structure recovery rates on simulated benchmarks, indicating that the model learns a spectrum-to-structure mapping precise enough to differentiate between polymorphs. To address generalization to experimental data, we compare a full-spectrum encoding against an encoding based on peak descriptors. The peak-based encoding generalizes substantially better, outperforming even a model trained on full spectra with augmentations fitted to the experimental noise distribution. These results demonstrate that representations robust to the noise and artifacts present in real-world PXRD offer a practical and scalable path toward closing the simulation-to-experiment gap, enabling zero-shot crystal structure solution from experimental PXRD with full or partial chemical composition input.

cond-mat.mtrl-sci↗
arXiv2026

MetaDNS: Enhancing Exploration in Discrete Neural Samplers via Well-Tempered Metadynamics

Sampling from discrete distributions with multiple modes and energy barriers is fundamental to machine learning and computational physics. Recent discrete neural samplers like MDNS suffer from mode collapse and fail to sample high-energy barrier regions between modes, which is critical for free energy estimation and understanding phase transitions. We propose Metadynamics Discrete Neural Sampler (MetaDNS), a general framework integrating well-tempered metadynamics into discrete diffusion or autoregressive samplers. By maintaining an adaptive, history-dependent bias potential along selected low-dimensional coordinates, MetaDNS forces exploration of previously inaccessible regions, enabling free energy reconstruction infeasible with standard neural samplers due to a lack of high-energy samples. On challenging low-temperature benchmarks including Ising, Potts, and the copper-gold binary alloy, MetaDNS reproduces the thermodynamic distribution. Compared to MCMC-based metadynamics, MetaDNS also achieves comparable exploration requiring fewer bias deposition steps.

cond-mat.stat-mech↗
arXiv2026

Harnessing AtomisticSkills for Agentic Atomistic Research

Computational materials science and chemistry span vast knowledge domains and fractured software ecosystems. Although large language models (LLMs) have demonstrated research capabilities, scaling monolithic agents to manage the rigor and complexity of atomistic research remains a challenge. Here, we introduce AtomisticSkills, an open-source harness framework that empowers general-purpose AI coding agents to conduct atomistic research across materials science, chemistry, and drug discovery. By hierarchically decomposing scientific workflows into agent skills and tools, AtomisticSkills provides agents with modular, extensible, and plug-and-play research capabilities. The framework integrates more than 100 human-curated multidisciplinary skills, including database access, thermodynamics and kinetics modeling, and diverse simulation engines employing machine learning interatomic potentials (MLIPs) and density functional theory (DFT). We validate its functional coverage against scientific literature and demonstrate robust orchestration capabilities across diverse scientific campaigns: generative design of Li-ion solid-state electrolytes, high-throughput screening of metal-organic frameworks for CO2 capture, autonomous MLIP benchmarking and fine-tuning, multi-stage structure-based virtual screening for drug design, multimodal X-ray diffraction pattern analysis, and screening of Fe-oxide catalysts for oxygen evolution reaction. AtomisticSkills provides a critical agent infrastructure towards building fully autonomous AI scientists.

physics.chem-ph↗
arXiv2026

Reweighting free energy profiles between universal machine learning interatomic potentials for fast consensus building

Free energy profiles serve as a fundamental bridge between microscopic atomic fluctuations and macroscopic thermodynamic observables. Estimating the free energy profile along a reaction coordinate, referred to as the potential of mean force (PMF), with density functional theory (DFT) accuracy is computationally expensive. Universal machine learning interatomic potentials (MLIPs) drastically reduce this cost, but their accuracy is strongly determined by their training data and hence can be uncertain for a given system. In this work, we present a systematic and scalable framework for reweighting PMFs, initially sampled with a single 'source' MLIP, across a representative suite of target MLIPs. Because traditional direct exponential reweighting fails for large system sizes due to low phase-space overlap between potentials, we deploy robust analytical corrections. Applying this to a complex 601-atom system of Li$^+$ transport in a nanoconfined electrolyte, we demonstrate that a mean energy-gap approximation effectively bypasses statistical collapse, producing a highly stable PMF matching the target PMF. Using this approach, we recover high-fidelity target thermodynamics across multiple DFT reference levels (PBE+D3, PBE-sol, r$^2$SCAN,r$^2$SCAN-D4) at a fraction of the computational cost of full simulations. Furthermore, thermodynamic analysis reveals that the studied MLIPs partition into two distinct clusters driven by their training data. Our reweighting framework successfully recovers target thermodynamic properties--specifically, reaction and activation free energies--even when the phase-space overlap between potentials is critically low. Ultimately, this approach establishes a vital diagnostic protocol to achieve affordable cross-model consensus on materials chemistry properties without redundant, resource-intensive simulations.

physics.chem-ph↗
arXiv2026

A Priori Sampling of Transition States with Guided Diffusion

Transition states, the first-order saddle points on the potential energy surfaces, govern the kinetics and mechanisms of chemical reactions and conformational changes. Locating them is challenging because transition pathways are topologically complex and can proceed via an ensemble of diverse routes. Existing methods address these challenges by introducing heuristic assumptions about the pathway or reaction coordinates, which limits their applicability when a good initial guess is unavailable or when the guess precludes alternative, potentially relevant pathways. We propose to bypass such heuristic limitations by introducing ASTRA, A Priori Sampling of TRAnsition States with Guided Diffusion, which reframes the transition state search as an inference-time scaling problem for generative models. ASTRA trains a score-based diffusion model on configurations from known metastable states. Then, ASTRA guides inference toward the isodensity surface separating the basins of metastable states via a principled composition of conditional scores. A Score-Aligned Ascent (SAA) process then approximates a reaction coordinate from the difference between conditioned scores and combines it with physical forces to drive convergence onto first-order transition states. Validated on benchmarks ranging from 2D potentials to biomolecular conformational changes and a chemical reaction, ASTRA locates transition states with high precision and discovers multiple reaction pathways, enabling mechanistic studies of complex molecular systems.

physics.chem-ph↗
arXiv2026

Computational Design and Experimental Validation of Photoactive PARP1 Inhibitors

Light-activated drugs are a promising way to treat localized diseases for which existing treatments have severe side effects. However, their development is complicated by the set of photophysical and biological properties that must be simultaneously optimized. Here we used computational techniques to find a set of promising candidates for the photoactive inhibition of the poly(ADP-ribose) polymerase 1 (PARP1) cancer target. Using our recently developed methods based on atomistic simulation and machine learning (ML), we screened a set of 5 million hypothetical photoactive ligands. Our workflow used protein-ligand docking to identify candidates with differential PARP1 binding under light and dark conditions; ML force fields and quantum chemistry calculations to predict p$K_\mathrm{a}$, absorption spectra, and thermal half-lives; graph-based surrogate models to screen additional compounds; excited-state nonadiabatic dynamics with ML force fields to estimate quantum yields; and free energy perturbation (FEP) to refine binding predictions. From these predictions, we prioritized a small set of synthetically feasible candidates expected to have red-shifted absorption spectra, thermal half-lives on the order of seconds to minutes, and isomer-dependent PARP1 binding under visible-light control. We synthesized 10 candidates and experimentally characterized their photobehavior and PARP1 inhibition constants. Among the validated compounds, \textbf{1} showed a 15-fold increase in inhibition of PARP1 upon green-light irradiation at 519 nm (208.8 $\pm$ 28.3 $μ$M vs 14.4 $\pm$ 1.9 $μ$M). These results validate the computation-guided screening strategy for identifying red-shifted PARP1 photoinhibitors, while also underscoring current limitations such as rapid thermal relaxation in aqueous media.

physics.chem-ph↗
arXiv2026

Scaling Autoregressive Models for Lattice Thermodynamics

Predicting how materials behave under realistic conditions requires understanding the statistical distribution of atomic configurations on crystal lattices, a problem central to alloy design, catalysis, and the study of phase transitions. Traditional Markov-chain Monte Carlo sampling suffers from slow convergence and critical slowing down near phase transitions, motivating the use of generative models that directly learn the thermodynamic distribution. Existing autoregressive models (ARMs), however, generate configurations in a fixed sequential order and incur high memory and training costs, limiting their applicability to realistic systems. Here, we develop a framework combining any-order ARMs, which generate configurations flexibly by conditioning on any known subset of lattice sites, with marginalization models (MAMs), which approximate the probability of any partial configuration in a single forward pass and substantially reduce memory requirements. This combination enables models trained on smaller lattices to be reused for sampling larger systems, while supporting expressive Transformer architectures with lattice-aware positional encodings at manageable computational cost. We demonstrate that Transformer-based any-order MAMs achieve more accurate free energies than multilayer perceptron-based ARMs on both the two-dimensional Ising model and CuAu alloys, faithfully capturing phase transitions and critical behavior. Overall, our framework scales from $10 \times 10$ to $20 \times 20$ Ising systems and from $2 \times 2 \times 4$ to $4 \times 4 \times 8$ CuAu supercells at reduced computational cost compared to conventional sampling methods.

cond-mat.stat-mech↗
arXiv2026

Learning Collective Variables from BioEmu with Time-Lagged Generation

Molecular dynamics is crucial for understanding molecular systems but its applicability is often limited by the vast timescales of rare events like protein folding. Enhanced sampling techniques overcome this by accelerating the simulation along key reaction pathways, which are defined by collective variables (CVs). However, identifying effective CVs that capture the slow, macroscopic dynamics of a system remains a major bottleneck. This work proposes a novel framework coined BioEmu-CV that learns these essential CVs automatically from BioEmu, a recently proposed foundation model for generating protein equilibrium samples. In particular, we re-purpose BioEmu to learn time-lagged generation conditioned on the learned CV, i.e., predict the distribution of molecular states after a certain amount of time. This training process promotes the CV to encode only the slow, long-term information while disregarding fast, random fluctuations. We validate our learned CV on fast-folding proteins with two key applications: (1) estimating free energy differences using on-the-fly probability enhanced sampling and (2) sampling transition paths with steered molecular dynamics. Our empirical study also serves as a new systematic and comprehensive benchmark for MLCVs on fast-folding proteins larger than Alanine Dipeptide.

cs.LG↗
arXiv2026

Discovery of Polymer Electrolytes with Bayesian Optimization and High-Throughput Molecular Dynamics simulations

Polymer electrolytes are critical for safe, high-energy-density solid-state batteries, yet discovering candidates that balance high ionic conductivity with high transference numbers remains a significant challenge. In this work, we develop a high-throughput screening platform that utilizes molecular dynamics simulations to navigate a chemical space of 1.7 million hypothetical polymer electrolyte candidates. Data from previous literature is used to warm-start batch Bayesian optimization for iteratively selecting new polymer electrolytes to evaluate. We iteratively identified, evaluated and analyzed 767 homopolymers as potential candidates. Our results reveal several candidates with transport properties exceeding the benchmark polyethylene oxide (PEO)/LiTFSI system. Crucially, our optimization campaigns for ionic conductivity and Li-diffusivity demonstrate that branched architectures and ketone functional groups significantly enhance ion-hopping mechanisms within the polymer matrix. We provide an in-depth mechanistic comparison of Li vs. Na cation transport and offer our open-source framework to accelerate the discovery of liquid, gel, and multi-cation electrolyte systems.

cond-mat.mtrl-sci↗