Explainable graph neural networks for organic cages
Kim Jelfs

June 23, 2022, 1 p.m.

Google Scholar Linkedin Twitter Twitter


The development of accurate and explicable machine learning models to predict the properties of topologically complex systems is a challenge in materials science. Porous organic cages, a class of polycyclic molecular materials, have potential application in molecular separations, catalysis and encapsulation. For most applications of porous organic cages, having a permanent internal cavity in the absence of solvent, a property termed “shape persistence” is critical. Here, we report the development of Graph Neural Networks (GNNs) to predict the shape persistence of organic cages. Graph neural networks are a class of neural networks where the data, in our case that of organic cages, are represented by graphs.

Brief CV

Kim Jelfs completed her PhD in Computational Chemistry at University College London, working on the development and application of modelling to understand zeolite crystal growth and was awarded the Ramsay Medal for the best completing PhD student. She was subsequently a visiting researcher at the Universitat de Barcelona, working with Prof. Stefan Bromley on the prediction of silicate cluster formation, before moving to the University of Liverpool, working as a PDRA across the experimental groups of Profs. Matt Rosseinsky and Andy Cooper. She was focused upon modelling porous materials, with my expertise spanning zeolites, metal-organic frameworks, polymers and porous molecular materials. Several of her computational predictions were experimentally realised. Since 2013 she has held a Royal Society University Research Fellowship (URF) at Imperial College London, allowing her to establish an independent research group. Her fellowship is entitled “Directing the synthesis of functional molecular materials”.



Share
Explainable graph neural networks for organic cages
Kim Jelfs

June 23, 2022, 1 p.m.

Google Scholar Linkedin Twitter Twitter


The development of accurate and explicable machine learning models to predict the properties of topologically complex systems is a challenge in materials science. Porous organic cages, a class of polycyclic molecular materials, have potential application in molecular separations, catalysis and encapsulation. For most applications of porous organic cages, having a permanent internal cavity in the absence of solvent, a property termed “shape persistence” is critical. Here, we report the development of Graph Neural Networks (GNNs) to predict the shape persistence of organic cages. Graph neural networks are a class of neural networks where the data, in our case that of organic cages, are represented by graphs.

Brief CV

Kim Jelfs completed her PhD in Computational Chemistry at University College London, working on the development and application of modelling to understand zeolite crystal growth and was awarded the Ramsay Medal for the best completing PhD student. She was subsequently a visiting researcher at the Universitat de Barcelona, working with Prof. Stefan Bromley on the prediction of silicate cluster formation, before moving to the University of Liverpool, working as a PDRA across the experimental groups of Profs. Matt Rosseinsky and Andy Cooper. She was focused upon modelling porous materials, with my expertise spanning zeolites, metal-organic frameworks, polymers and porous molecular materials. Several of her computational predictions were experimentally realised. Since 2013 she has held a Royal Society University Research Fellowship (URF) at Imperial College London, allowing her to establish an independent research group. Her fellowship is entitled “Directing the synthesis of functional molecular materials”.



Share