A5067139007- Machine Learning in Materials Science
- Protein Structure and Dynamics
- Advanced Graph Neural Networks
- Computational Drug Discovery Methods
- Spectroscopy and Quantum Chemical Studies
- Crystallography and molecular interactions
- Enzyme Structure and Function
- Advanced Chemical Physics Studies
- Graph Theory and Algorithms
- Complex Network Analysis Techniques
- Advanced Clustering Algorithms Research
University College London
2024
University of Southampton
2021-2023
The ensemble of structures generated by molecular mechanics (MM) simulations is determined the functional form force field employed and its parameterization. For a given form, quality parameterization crucial will determine how accurately we can compute observable properties from simulations. While accurate parameterizations are available for biomolecules, such as proteins or DNA, new molecules, drug candidates, particularly challenging these may involve groups interactions which parameters...
We present a comparative study that evaluates the performance of machine learning potential (ANI-2x), conventional force field (GAFF), and an optimally tuned GAFF-like in modeling set 10 γ-fluorohydrins exhibit complex interplay between intra- intermolecular interactions determining conformer stability. To benchmark each molecular model, we evaluated their energetic, geometric, sampling accuracies relative to quantum-mechanical data. This involved conformational analysis both gas phase...
The recent trend in using network and graph structures to represent a variety of different data types has renewed interest the partitioning (GP) problem. This stems from need for general methods that can both efficiently identify communities reduce dimensionality large graphs while satisfying various application-specific criteria. Traditional clustering algorithms often struggle capture complex relationships within generalize arbitrary emergence neural networks (GNNs) as powerful framework...
Conformational analysis is of paramount importance in drug design: it crucial to determine pharmacological properties, understand molecular recognition processes, and characterize the conformations ligands when unbound. Molecular Mechanics (MM) simulation methods, such as Monte Carlo (MC) dynamics (MD), are usually employed generate ensembles structures due their ability extensively sample conformational space molecules. The accuracy these MM-based schemes strongly depends on functional form...
Traditional clustering algorithms often struggle to capture the complex relationships within graphs and generalise arbitrary criteria. The emergence of graph neural networks (GNNs) as a powerful framework for learning representations data provides new approaches solving problem. Previous work has shown GNNs be capable proposing partitionings using variety criteria, however, these have not yet been extended on Markov chains or kinetic networks. These arise frequently in study molecular...