Closing date: 19 July 2023
Full-time, fixed-term contract for 36 months.
We are seeking a theoretical Postdoctoral Research Fellow (PDRF) to employ generative modelling and deep learning to infer mathematical models of cellular force sensing and response. Experimental images of cell-cell interactions will be acquired through state-of-the-art 3D lightsheet microscopy and employing novel microfluidics chambers. This is an EPSRC / NSF funded project in close collaboration with Warwick Computer Science, Warwick Medical School and Duquesne University (grant number EP/X026663/1) .
You will be joining a dynamic Bioimage Informatics and Computational Biology group in the Department of Computer Science, University of Warwick, which has a track record in image analysis, applied machine learning and computational modelling of cellular dynamics. We are based in the Mathematical Sciences Building, which is shared between the departments of Computer Science, Statistics and Mathematics, a unique environment to interact with top researchers in computational sciences.
The focus of the project will be on developing a new technology platform combining light sheet microscopy, microfluidics, omics and computational image analysis and modelling to answer outstanding questions in mechanobiology. The project will bring together experts in cell and developmental biology, biophysics, engineering, computer science and bioinformatics. We aim to identify new molecular and cellular determinants and pathways controlling mechanical cell-cell communication using a stem cell model.
Our international research team will be joined by a second, experimental PDRA in the lab of Dr Michael Smutny at Warwick Medical School who will conduct molecular biology experiments and 3D microscopy. Another researcher will be joining the lab of Dr Melikhan Tanyeri at the University of Pittsburgh, Department of Engineering (US).
Your work will build on our recent advances in 3D segmentation , Virtual Reality solutions for annotating 3D data (MiCellAnnGELo) , 3D generative models of microscopy data , and graph neural network-based methods , which we are currently applying to dynamic cell features. The first goal will be to employ generative modelling techniques (generative adversarial networks) to combine two-channel image data from experiments with one fluorescent reference marker and additional biomarkers of interest, to capture and map the complex spatio-temporal dynamics of a larger set of different biomarkers involved in force sensing and its response. The second goal will be to develop predictive mathematical models that can inform our biological understanding of force sensing. Depending on your background, this goal will either employ modern neural network-based models for dynamical systems, or more traditional PDE-based approaches in combination with statistical techniques for parameter estimation.
Suitable candidates will have at least a 2.1 Honours degree and a PhD in Mathematics, Data or Computer Science, or Physics. You will have a strong background in one or more of the following: machine learning; mathematical/biophysical modelling; computer vision.
Earliest start date is 1st September.
For informal enquiries please get in touch with Till.Bretschneider@warwick.ac.uk.