New book published!
January 22, 2024
MetaSimulation of NonEquilibrium Processes
The MSNEP Group
The book builds on the analogy between social groups and assemblies of molecules to introduce the concepts of statistical mechanics, machine learning and data science. Applying a data analytics approach to molecular systems, we show how individual (molecular) features and interactions between molecules, or "communication" processes, allow for the prediction of properties and collective behavior of molecular systems - just as polling and social networking shed light on the behavior of social groups. Applications to cutting-edge research for biological, environmental, and energy systems are also presented.
Developing new pharmaceutical compounds is a lengthy, costly, and intensive process. In recent years, the development of Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) models has drawn considerable interest in drug discovery. In this review, we discuss recent advances in the field and show how these methods can be leveraged to assist each stage of the drug discovery process.
The ability of active matter to assemble into reconfigurable nonequilibrium structures has drawn considerable interest in recent years. We investigate how active fluids respond to spatial light patterns through simulations and experiments on light-activated self-propelled colloidal particles. We examine the processes of inverse templated assembly, which involves creating a region without active particles through a bright pattern, and templated assembly, which promotes the formation of dense particle regions through a dark pattern. We identify scaling relations for the characteristic times for both processes that quantify the interplay between the dimension of the applied pattern and the intrinsic properties of the active fluid.
The past years have been marked by fast-paced advances in the data-driven and Machine Learning (ML)-aided design and characterization of energy materials. Equally remarkable is the broad range of theoretical methods, physical systems, and applications that ML-aided research has enabled to explore. This symposium will bring together chemists, physicists, and engineers from these different areas, with complementary computational and experimental expertise, to share the cutting-edge advances in the design of energy materials enabled by ML in their respective fields. Join us for this exciting symposium co-organized with Mingda Li (MIT) and Fang Liu (Emory)
Understanding and optimizing solid-solid interfaces in graphene-based nanocomposite catalysts is crucial for electrocatalytic applications such as CO2 reduction. Here we develop a new strategy for constructing innovative electrocatalysts from metal nanoclusters covalently attached onto three-dimensional pristine graphene electrodes through rationally designed molecular linkers. Using a combination of synthesis, computational modeling, electrochemistry, and spectroscopy, we show how the control over solid-solid interfaces impacts performance in electrochemical devices. Stay tuned for more news and work with our collaborators Kwok-Fan Chow and Mingdi Yan at UML and Gonghu Li at UNH.
Read the paper here. A modulated fingerprint assisted machine learning method for retrieving elastic moduli from resonant ultrasound spectroscopy. We used deep-learning-based models to automatically obtain elastic moduli from resonant ultrasound spectroscopy (RUS) spectra, which conventionally require user intervention of published analysis codes. By strategically converting theoretical RUS spectra into their modulated fingerprints and using them as a dataset to train neural network models, we obtained models that successfully predicted elastic moduli. In summary, our modulated fingerprint method is an efficient tool to transform raw spectroscopy data and train neural network models with high accuracy and resistance to spectra distortion.
Living systems have the unique ability to form hierarchical assemblies, in which individual constituents can perform tasks cooperatively and emergently. Harnessing such properties is a long-standing challenge for the rational design of dynamic materials, that can respond to their environment, communicate with one another, and undergo a rapid, reversible, assembly through the transduction of energy. Here we develop a combined experimental, computational, theoretical and Machine Learning framework to program the assembly of smart active materials. Stay tuned for more news and work with our collaborators Paul Chaikin, Stefano Sacanna and Mark Tuckerman at NYU.
NSF award on ML for molecular crystals
September 1, 2022
Molecular crystals play an essential role in the pharmaceutical, agrochemical, electronics, and defense industries. In many instances, a given chemical compound may have more than one crystal structure, a phenomenon known as polymorphism. A crystal may also contain impurities, the most important among these being water. Such structures are referred to as crystal hydrates. The ability of these materials to function in a desired manner may depend on which structure, pure or impure, they form. Utilizing advances in high-performance computing and artificial intelligence, we create new computational approaches and software components for rapidly predicting polymorphic structures in molecular crystals and understanding the transitions between structures. More info about our latest work with our collaborator Mark Tuckerman at NYU very soon.