arXiv subjects

Kevin Ryczko

Publications and source records attributed to Kevin Ryczko.

14 recordsLinked to original sources

arXiv2025

AQCat25: Unlocking spin-aware, high-fidelity machine learning potentials for heterogeneous catalysis

Large-scale datasets have enabled highly accurate machine learning interatomic potentials (MLIPs) for general-purpose heterogeneous catalysis modeling. There are, however, some limitations in what can be treated with these potentials because of gaps in the underlying training data. To extend these capabilities, we introduce AQCat25, a complementary dataset of 13.5 million density functional theory (DFT) single point calculations designed to improve the treatment of systems where spin polarization and/or higher fidelity are critical. We also investigate methodologies for integrating new datasets, such as AQCat25, with the broader Open Catalyst 2020 (OC20) dataset to create spin-aware models without sacrificing generalizability. We find that directly tuning a general model on AQCat25 leads to catastrophic forgetting of the original dataset's knowledge. Conversely, joint training strategies prove effective for improving accuracy on the new data without sacrificing general performance. This joint approach introduces a challenge, as the model must learn from a dataset containing both mixed-fidelity calculations and mixed-physics (spin-polarized vs. unpolarized). We show that explicitly conditioning the model on this system-specific metadata, for example by using Feature-wise Linear Modulation (FiLM), successfully addresses this challenge and further enhances model accuracy. Ultimately, our work establishes an effective protocol for bridging DFT fidelity domains to advance the predictive power of foundational models in catalysis.

cond-mat.mtrl-sci↗

arXiv2024

Guided Multi-objective Generative AI to Enhance Structure-based Drug Design

Generative AI has the potential to revolutionize drug discovery. Yet, despite recent advances in deep learning, existing models cannot generate molecules that satisfy all desired physicochemical properties. Herein, we describe IDOLpro, a generative chemistry AI combining diffusion with multi-objective optimization for structure-based drug design. Differentiable scoring functions guide the latent variables of the diffusion model to explore uncharted chemical space and generate novel ligands in silico, optimizing a plurality of target physicochemical properties. We demonstrate our platform's effectiveness by generating ligands with optimized binding affinity and synthetic accessibility on two benchmark sets. IDOLpro produces ligands with binding affinities over 10%-20% better than the next best state-of-the-art method on each test set, producing more drug-like molecules with generally better synthetic accessibility scores than other methods. We do a head-to-head comparison of IDOLpro against a classic virtual screen of a large database of drug-like molecules. We show that IDOLpro can generate molecules for a range of important disease-related targets with better binding affinity and synthetic accessibility than any molecule found in the virtual screen while being over 100x faster and less expensive to run. On a test set of experimental complexes, IDOLpro is the first to produce molecules with better binding affinities than experimentally observed ligands. IDOLpro can accommodate other scoring functions (e.g. ADME-Tox) to accelerate hit-finding, hit-to-lead, and lead optimization for drug discovery.

physics.chem-ph↗

arXiv2023

Accelerated Organic Crystal Structure Prediction with Genetic Algorithms and Machine Learning

We present a high-throughput, end-to-end pipeline for organic crystal structure prediction (CSP) -- the problem of identifying the stable crystal structures that will form from a given molecule based only on its molecular composition. Our tool uses Neural Network Potentials (NNPs) to allow for efficient screening and structural relaxations of generated crystal candidates. Our pipeline consists of two distinct stages -- random search, whereby crystal candidates are randomly generated and screened, and optimization, where a genetic algorithm (GA) optimizes this screened population. We assess the performance of each stage of our pipeline on 21 molecules taken from the Cambridge Crystallographic Data Centre's CSP blind tests. We show that random search alone yields matches for $\approx 50\%$ of targets. We then validate the potential of our full pipeline, making use of the GA to optimize the Root Mean-Squared Deviation (RMSD) between crystal candidates and the experimentally derived structure. With this approach, we are able to find matches for $\approx80\%$ of candidates with 10-100 times smaller initial population sizes than when using random search. Lastly, we run our full pipeline with an ANI model that is trained on a small dataset of molecules extracted from crystal structures in the Cambridge Structural Database, generating $\approx 60\%$ of targets. By leveraging ML models trained to predict energies at the DFT level, our pipeline has the potential to approach the accuracy of \emph{ab initio} methods and the efficiency of empirical force-fields.

cond-mat.mtrl-sci↗

arXiv2022

Machine Learning Diffusion Monte Carlo Energies

We present two machine learning methodologies that are capable of predicting diffusion Monte Carlo (DMC) energies with small datasets (~60 DMC calculations in total). The first uses voxel deep neural networks (VDNNs) to predict DMC energy densities using Kohn-Sham density functional theory (DFT) electron densities as input. The second uses kernel ridge regression (KRR) to predict atomic contributions to the DMC total energy using atomic environment vectors as input (we used atom centred symmetry functions, atomic environment vectors from the ANI models, and smooth overlap of atomic positions). We first compare the methodologies on pristine graphene lattices, where we find the KRR methodology performs best in comparison to gradient boosted decision trees, random forest, gaussian process regression, and multilayer perceptrons. In addition, KRR outperforms VDNNs by an order of magnitude. Afterwards, we study the generalizability of KRR to predict the energy barrier associated with a Stone-Wales defect. Lastly, we move from 2D to 3D materials and use KRR to predict total energies of liquid water. In all cases, we find that the KRR models are more accurate than Kohn-Sham DFT and all mean absolute errors are less than chemical accuracy.

cond-mat.mes-hall↗

arXiv2022

Toward Orbital-Free Density Functional Theory with Small Data Sets and Deep Learning

We use voxel deep neural networks to predict energy densities and functional derivatives of electron kinetic energies for the Thomas-Fermi model and Kohn-Sham density functional theory calculations. We show that the ground-state electron density can be found via direct minimization for a graphene lattice without any projection scheme using a voxel deep neural network trained with the Thomas-Fermi model. Additionally, we predict the kinetic energy of a graphene lattice within chemical accuracy after training from only 2 Kohn-Sham density functional theory (DFT) calculations. We identify an important sampling issue inherent in Kohn-Sham DFT calculations and propose future work to rectify this problem. Furthermore, we demonstrate an alternative, functional derivative-free, Monte Carlo based orbital free density functional theory algorithm to calculate an accurate 2-electron density in a double inverted Gaussian potential with a machine-learned kinetic energy functional.

cond-mat.mes-hall↗

arXiv2021

Electronic Response Quantities of Solids and Deep Learning

We introduce a deep neural network (DNN) framework called the \textbf{r}eal-space \textbf{a}tomic \textbf{d}ecomposition \textbf{net}work (\textsc{radnet}), which is capable of making accurate polarization and static dielectric function predictions for solids. We use these predictions to calculate Born-effective charges, longitudinal optical transverse optical (LO-TO) splitting frequencies, and Raman tensors for two prototypical examples: GaAs and BN. We then compute the Raman spectra, and find excellent agreement with \textit{ab initio} techniques. \textsc{radnet} is as good or better than current methodologies. Lastly, we discuss how \textsc{radnet} scales to larger systems, paving the way for predictions of response functions on meso-scale structures with \textit{ab initio} accuracy.

cond-mat.mes-hall↗

arXiv2021

Neural evolution structure generation: High Entropy Alloys

We propose a method of neural evolution structures (NESs) combining artificial neural networks (ANNs) and evolutionary algorithms (EAs) to generate High Entropy Alloys (HEAs) structures. Our inverse design approach is based on pair distribution functions and atomic properties and allows one to train a model on smaller unit cells and then generate a larger cell. With a speed-up factor of approximately 1000 with respect to the SQSs, the NESs dramatically reduces computational costs and time, making possible the generation of very large structures (over 40,000 atoms) in few hours. Additionally, unlike the SQSs, the same model can be used to generate multiple structures with the same fractional composition.

cond-mat.dis-nn↗

arXiv2021

Inverse Design of a Graphene-Based Quantum Transducer via Neuroevolution

We introduce an inverse design framework based on artificial neural networks, genetic algorithms, and tight-binding calculations, capable to optimize the very large configuration space of nanoelectronic devices. Our non-linear optimization procedure operates on trial Hamiltonians through superoperators controlling growth policies of regions of distinct doping. We demonstrate that our algorithm optimizes the doping of graphene-based three-terminal devices for valleytronics applications, monotonously converging to synthesizable devices with high merit functions in a few thousand evaluations (out of $\simeq 2^{3800}$ possible configurations). The best-performing device allowed for a terminal-specific separation of valley currents with $\simeq 96$\% ($\simeq 94\%)$ $K$ ($K'$) valley purity. Importantly, the devices found through our non-linear optimization procedure have both higher merit function and higher robustness to defects than the ones obtained through geometry optimization.

cond-mat.mes-hall↗

arXiv2021

Deep Learning and Density Functional Theory

We show that deep neural networks can be integrated into, or fully replace, the Kohn-Sham density functional theory scheme for multi-electron systems in simple harmonic oscillator and random external potentials with no feature engineering. We first show that self-consistent charge densities calculated with different exchange-correlation functionals can be used as input to an extensive deep neural network to make predictions for correlation, exchange, external, kinetic and total energies simultaneously. Additionally, we show that one can also make all of the same predictions with the external potential rather than the self-consistent charge density, which allows one to circumvent the Kohn-Sham scheme altogether. We then show that a self-consistent charge density found from a non-local exchange-correlation functional can be used to make energy predictions for a semi-local exchange-correlation functional. Lastly, we use a deep convolutional inverse graphics network to predict the charge density given an external potential for different exchange-correlation functionals and assess the viability of the predicted charge densities. This work shows that extensive deep neural networks are generalizable and transferable given the variability of the potentials (maximum total energy range $\approx100$ Ha), because they require no feature engineering, and because they can scale to an arbitrary system size with an $\mathcal{O}(N)$ computational cost.

cond-mat.mtrl-sci↗

arXiv2020

Twin Neural Network Regression

We introduce twin neural network (TNN) regression. This method predicts differences between the target values of two different data points rather than the targets themselves. The solution of a traditional regression problem is then obtained by averaging over an ensemble of all predicted differences between the targets of an unseen data point and all training data points. Whereas ensembles are normally costly to produce, TNN regression intrinsically creates an ensemble of predictions of twice the size of the training set while only training a single neural network. Since ensembles have been shown to be more accurate than single models this property naturally transfers to TNN regression. We show that TNNs are able to compete or yield more accurate predictions for different data sets, compared to other state-of-the-art methods. Furthermore, TNN regression is constrained by self-consistency conditions. We find that the violation of these conditions provides an estimate for the prediction uncertainty.

cs.LG↗

arXiv2019

Extensive deep neural networks for transferring small scale learning to large scale systems

We present a physically-motivated topology of a deep neural network that can efficiently infer extensive parameters (such as energy, entropy, or number of particles) of arbitrarily large systems, doing so with O(N) scaling. We use a form of domain decomposition for training and inference, where each sub-domain (tile) is comprised of a non-overlapping focus region surrounded by an overlapping context region. The size of these regions is motivated by the physical interaction length scales of the problem. We demonstrate the application of EDNNs to three physical systems: the Ising model and two hexagonal/graphene-like datasets. In the latter, an EDNN was able to make total energy predictions of a 60 atoms system, with comparable accuracy to density functional theory (DFT), in 57 milliseconds. Additionally EDNNs are well suited for massively parallel evaluation, as no communication is necessary during neural network evaluation. We demonstrate that EDNNs can be used to make an energy prediction of a two-dimensional 35.2 million atom system, over 1 square micrometer of material, at an accuracy comparable to DFT, in under 25 minutes. Such a system exists on a length scale visible with optical microscopy and larger than some living organisms.

physics.comp-ph↗

arXiv2018

Convolutional neural networks for atomistic systems

We introduce a new method, called CNNAS (convolutional neural networks for atomistic systems), for calculating the total energy of atomic systems which rivals the computational cost of empirical potentials while maintaining the accuracy of \emph{ab initio} calculations. This method uses deep convolutional neural networks (CNNs), where the input to these networks are simple representations of the atomic structure. We use this approach to predict energies obtained using density functional theory (DFT) for 2D hexagonal lattices of various types. Using a dataset consisting of graphene, hexagonal boron nitride (hBN), and graphene-hBN heterostructures, with and without defects, we trained a deep CNN that is capable of predicting DFT energies to an extremely high accuracy, with a mean absolute error (MAE) of 0.198 meV / atom (maximum absolute error of 16.1 meV / atom). To explore our new methodology, we investigate the ability of a deep neural network (DNN) in predicting a Lennard-Jones energy and separation distance for a dataset of dimer molecules in both two and three dimensions. In addition, we systematically investigate the flexibility of the deep learning models by performing interpolation and extrapolation tests.

cond-mat.mtrl-sci↗

arXiv2017

Structural characterizations of water-metal interfaces

We analyze and compare the structural, dynamical, and electronic properties of liquid water next to prototypical metals including Pt, graphite, and graphene. Our results are built on Born-Oppenheimer molecular dynamics (BOMD) generated using density functional theory (DFT) which explicitly include van der Waals (vdW) interactions within a first principles approach. All calculations reported use large simulation cells, allowing for an accurate treatment of the water-electrode interfaces. We have included vdW interactions through the use of the optB86b-vdW exchange correlation functional. Comparisons with the Perdew-Burke-Ernzerhof (PBE) exchange correlation functional are also shown. We find an initial peak, due to chemisorption, in the density profile of the liquid water-Pt interface not seen in the liquid water-graphite interface, liquid water-graphene interface, nor interfaces studied previously. To further investigate this chemisorption peak, we also report differences in the electronic structure of single water molecules on both Pt and graphite surfaces. We find that a covalent bond forms between the single water molecule and the Pt surface, but not between the single water molecule and the graphite surface. We also discuss the effects that defects and dopants in the graphite and graphene surfaces have on the structure and dynamics of liquid water.

cond-mat.mtrl-sci↗

arXiv2016

Hashkat: Large-scale simulations of online social networks

Hashkat (http://hashkat.org) is a free, open source, agent based simulation software package designed to simulate large-scale online social networks (e.g. Twitter, Facebook, LinkedIn, etc). It allows for dynamic agent generation, edge creation, and information propagation. The purpose of hashkat is to study the growth of online social networks and how information flows within them. Like real life online social networks, hashkat incorporates user relationships, information diffusion, and trending topics. Hashkat was implemented in C++, and was designed with extensibility in mind. The software includes Shell and Python scripts for easy installation and usability. In this report, we describe all of the algorithms and features integrated into hashkat before moving on to example use cases. In general, hashkat can be used to understand the underlying topology of social networks, validate sampling methods of such networks, develop business strategy for advertising on online social networks, and test new features of an online social network before going into production.

cs.SI↗