Skip to the content.

Data-Efficient and Interpretable AI for Reaction Prediction and Molecular Design

Project ID: 2531ad1500

(You will need this ID for your application)

Research Theme: Engineering

UCL Lead department: Chemical Engineering

Department Website

Lead Supervisor: Lauren Ye Seol Lee

Project Summary:

Project description: The accelerated discovery of novel molecules increasingly depends on predictive modelling in chemical synthesis. Advances in AI now allow for systematic predictions of reaction outcomes, minimizing traditional trial-and-error methods. Yet, current AI models often lack generalizability and interpretability due to narrow datasets and computationally intensive descriptors, which limits their applicability across diverse reactions. Addressing these challenges is crucial for advancing sustainable synthesis and efficient molecular discovery.

This project focuses on developing a reaction-centric AI foundational model that incorporates 3D molecular data, including critical steric and electronic descriptors, to improve model interpretability and predictive accuracy. You will work with self-supervised learning techniques with a special focus on graph neural networks, to capture essential structure-performance relationships and design a data-efficient framework that minimizes dependence on computationally intensive methods, such as density functional theory (DFT). Additionally, you will contribute to developing a computer-aided molecular design (CAMD) framework to support applications in medicinal chemistry, such as identifying optimal green solvents that enhance reaction kinetics (i.e., reaction rate).

Who you will be working with: This project is supervised by Dr Lauren Lee in collaboration with Prof Federico Galvanin at Sargent Centre for Process Systems Engineering at UCL, a world-leading research hub in chemical engineering. You will benefit from expert guidance in AI-driven chemical synthesis, process systems engineering, chemical engineering, gaining hands-on experience in advanced research methods. The position also offers opportunities to engage with global academic and industry experts, with potential access to state-of-the-art robotic chemical synthesis facilities at the Manufacturing Future Lab of UCL East.

Who we are looking for: We seek an enthusiastic, self-motivated candidate with a 2:1 MEng/MSc or equivalent in Chemical Engineering, Chemistry, or a related field. A solid background in mathematical modelling, optimization, computational chemistry, and/or data science is desirable, along with an interest in CAMD technologies.