The position is associated with the RESPOND3 project “Towards better computational approaches and responsible innovation strategies in early drug discovery: application to antibiotics and COPD”. RESPOND3 is part of the Centre for Digital Live Norway (https://digitallifenorway.org ) and is a cross-disciplinary collaboration between research groups at the Centre for Advanced studies in Biomedical Innovation Law at the University of Copenhagen, the Department of Computer Science, Electrical Engineering and Mathematical Sciences at the Western Norway University of Applied Sciences as well as the Department of Biomedicine and the Department of Chemistry at University of Bergen. The main focus of RESPOND3 will be hit and lead optimization for targets for antibiotics (SAM riboswitch) and chronic obstructive pulmonary disease (proteinase 3) guided by the use of computational methods as well as the development of new scoring functions using machine learning. The entire project is embedded in a responsible research and innovation strategy.
The RESPOND3 research team will eventually consist of four postdocs, three PhD students and five principle investigators, covering medicinal chemistry, computational chemistry and biology, structural biology, computer science, mathematics, and law. In addition to working with the RESPOND3 team, the postdoctoral fellow will be part of the vibrant research community at the recently established Mohn Medical Imaging and Visualization Center (MMIV, https://mmiv.no ), Haukeland University Hospital, with access to world-class computing infrastructure and to top expertise in life sciences, machine learning and computation. The researcher will also be associated with the ICT-oriented strategic research programme in computer science at HVL (http://ict.hvl.no) which spans the areas of software engineering, sensor networks, machine learning, and engineering computing.
The announced postdoc position will focus on applications of machine learning in drug discovery and development. The exact direction of research will depend on the background and interests of the successful candidate. Potential research areas and directions include
Development of new scoring functions for protein-ligand interactions using both in-house data and publicly available data sets
Attempts at combatting the tendency for scoring functions to exploit uninteresting features in the training data (data leakage) using e.g. various debiasing techniques. The improved scoring functions will be evaluated prospectively.
Investigation of the potential for machine learning methods in de novo structure generation (e.g. using variants of generative adversarial networks)
In addition, the successful candidate will
Assist in the supervision of MSc students and PhD candidates.
Contribute to the development of joint project applications for external funding in collaboration with permanent research staff members at HVL, MMIV and UiB