4 Year MRC PhD Programme* Quantitative phenotyping by machine learning

Distinguishing phenotypes in microscopy images involves complex pattern recognition. The quantification of phenotypes by computer-based image recognition enables machine learning to classify the variation in biological response1. Image classification by machine learning is an essential step in the development of automated, high-throughput screening of biological phenotypes. Furthermore, by integrating information on chemical structures, gene ontology and pathway information it is possible to develop an iterative (‘smart’) screening process that learns which chemical structures to screen next from a large library of possible agents. The project will focus on the development of machine learning methods to classify phenotypes in real time to development ‘smart’, adaptive, iterative screening methods2 for high-content screening.  Recent advances in large-scale image management for high-content screening data3 provides a framework for the large-scale developing quantitative phenotyping models…


Link to Full Article: 4 Year MRC PhD Programme* Quantitative phenotyping by machine learning