Data and Machine Intelligence

Chen Group: "Data-driven materials exploration and optimization"


We build computational capabilities at the intersection of chemistry, chemical engineering, and computer science to underpin atomic-scale materials design, by fusing chemical knowledge with state-of-the-art computation ranging from quantum mechanics to data-driven heuristics. Our current mission is to use computation to inform experiment, with a longer-term vision of synchronizing computation with experiment in closed-loop discovery and optimization of molecules and materials for functions.

Categories: Functional molecules and materials;optimization-driven design of experiments; explainable AI in chemistry

Albrecht Group: "Advanced data analysis and Deep Learning in single-molecule science"


We are using and developing tools for advanced single-molecule data analysis and unsupervised classification, including dimensionality reduction techniques, Convolutional Neural Networks and Transfer Learning approaches for applications in sensing and single-molecule electronics.

Categories: unsupervised classification; Transfer Learning; Deep Learning

Chakrabarti Group: "Machine learning soft materials"


We are interested in machine learning techniques for classifying colloidal crystal structures as well as recognising patterns in seemingly disordered soft materials susceptible to thermal fluctuations.

Categories: pattern recognition; classification; materials;

Gupta Group: "Optical sensing"


We are developing label-free optical methods and technologies for sensing of chemical and biological analytes. Hydrogels and biconjugation approaches are key to the development of these methods/ technologies. The raw output of these sensors is in the form of images, which are processed in real-time for continuous monitoring of the concentration of analytes. We are interested in exploring the use of AI to analyse the output of our sensors.

Categories: platforms; materials, data

Herten Group: "Fluorescent probes and single-molecule spectroscopy"


We are developing spectroscopy tools for quantitative and super-resolution fluorescence microscopy. Aiming at protein counting and 3D super-resolution representation of protein networks, we are highly interested to further improve interpretation of spectra and images and to support the design of new fluorescent probes by use of machine-learning approaches.

Categories: probes; platforms

Neely Group: " New tools for understanding genomes"


We develop novel chemistry that allow us to tag specific genomic sequences. We are using these chemistries in imaging DNA sequence, in nanopore sequencing experiments and as a way to better understand the epigenome. Our work creates large, complex but information-rich datasets and we employ AI to help us better understand this data.
Categories: probes; data.

Softley Group: "Chemical dynamics at ultracold temperatures"


We are developing new machine-learning approaches to the real-time analysis of fluorescence-images of Coulomb crystals – assemblies of 10s or 100s of laser-cooled ions in a Paul Trap that form ellipsoidal-shaped structures under the cooling/trapping conditions.  We use these crystals as a medium for studying the kinetics and dynamics of chemical reactions at ultracold temperatures, following reactions by observing changes in the Coulomb crystal structure.

Categories: imaging, deep learning, transfer learning

Styles Group: "Computational Chemical Imaging"


We work on understanding complex data from imaging experiments that directly probe sample chemistry, including labelled techniques such as fluorescence microscopy, and unlabelled techniques such as Raman and Mass Spectrometry Imaging.

Categories: imaging; probes; analytes