Machine learning discovery of post-transcriptionally regulated gene candidates using imaging and genomic data

The volume of microscopy images and their associated genome-wide metadata that are generated by many biologists is too large to be effectively browsed and interpreted in order to formulate novel testable hypotheses. Overcoming this challenge is of key importance for future biological discovery as data volumes continue to grow exponentially. The project will bridge this important technological gap in a unique partnership between a biomedical research lab lead by Prof. Ilan and Zegami (https://zegami.com), an Oxford University spinout company, whose founder and Chief Scientific Officer is Stephen Taylor. Zegami provides innovative cloud-based software to display vast databases of images sorted interactively in real time with complex metadata (example described in: https://www.youtube.com/watch?v=32bqn-Agt08). The main aim of the project is to develop machine learning software (using supervised random forest algorithms) that automatically generates scientific hypotheses based on correlations between existing high quality imaging data and genome-wide bioinformatics data, with guidance from the user. BBSRC REMIT AND FUNDING PORTFOLIO: "Tools and technology underpinning biological research", specifically "data driven biology" and "exploiting new ways of working" and the area of artificial intelligence using supervised machine learning algorithms.