AI-Driven Discovery
Extending doctoral research into Python-based cheminformatics and AI-assisted virtual screening for building the next layer of computational methodology for accelerated materials discovery.
Machine Learning & Python Libraries
Python-based tools supporting computational chemistry, cheminformatics, and AI-assisted materials discovery through molecular representation, descriptor generation, and computational workflow development.
RDKit
CheminformaticsAn open-source Python toolkit for cheminformatics used to process molecular structures, generate descriptors and fingerprints, perform substructure searches, and prepare molecular libraries for computational modelling and machine learning.
Applications
DeepChem
Learning & Research Interest
Exploring DeepChem for developing machine learning models that predict molecular and materials properties from computational descriptors, with the long-term goal of integrating AI into computational materials discovery workflows.
Research Interests
AI-Assisted Discovery of Functional Materials
A computational–experimental workflow that combines quantum chemical modelling, cheminformatics, and machine learning to accelerate the discovery of advanced functional materials for sensing, selective metal recovery, and sustainable technologies.
Descriptor Library
Generate a virtual library of metal–ligand systems and functional nanomaterials. Compute DFT-derived electronic descriptors, including HOMO–LUMO energies, electrostatic potential surfaces, charge distribution, and global reactivity descriptors.
Materials Property Screening
Rapidly screen candidates using physicochemical, structural, and electronic criteria to eliminate unstable or unsuitable materials before expensive calculations and experiments.
AI-Assisted Property Prediction
Apply machine learning models trained on computational and experimental data to predict structure–property relationships, selectivity, and material performance, enabling rapid prioritisation of the most promising candidates.
DFT Validation
Perform high-accuracy density functional theory calculations to validate the electronic structure, stability, and key properties of shortlisted materials before laboratory investigation.
Experimental Validation
Synthesize and characterize the highest-ranked candidates, using experimental results to refine computational models and continuously improve future predictions.
How AI Narrows Thousands of Materials to a Handful
Candidate Materials Library
Virtual library of metal–ligand systems and functional nanomaterials
Descriptor Generation
Generate DFT-derived electronic descriptors: HOMO–LUMO energies, electrostatic potential surfaces, charge distribution
AI Property Prediction
Machine learning models predict structure–property relationships, selectivity, stability, and functional performance
High-Fidelity DFT Validation
Perform high-accuracy DFT calculations to validate electronic structure, stability, and key material properties
Experimental Validation
Synthesize and characterize the highest-ranked candidates to verify computational predictions and generate new training data
Functional Materials
Validated materials for selective metal-ion sensing, critical metal recovery, and sustainable environmental applications
Each stage progressively narrows the search space, allowing computational intelligence and experimental validation to work together so laboratory effort is focused on the most promising functional materials.
Interested in Collaboration?
Open to collaborative research in AI-assisted materials discovery, cheminformatics, and computational–experimental workflows for advanced functional materials.