Visualising and Testing for Difference in High Throughput Bio-Imaging

Obtaining the sequence of numerous genomes and subsequent identification of the encoded proteome has created the need for large-scale systematic approaches to understand the functions of the tens of thousands of proteins at the cellular level. High-throughput automated fluorescent microscope imaging technologies enable the experimental determination of a protein’s sub-cellular localization and its dynamic trafficking. The desire and the ability to carry out high-throughput screenings of protein localization and trafficking for applications such as drug discovery is leading to a rapid growth in cell images in need of analysis on a scale comparable to that of the genomic revolution. However, the bottleneck in the pipeline is the need for a researcher to examine the wealth of data that modern automated microscopes are capturing. Much as in the genomic revolution, what is needed is a range of methodologies to search, classify, cluster and statistically compare large image sets to be able to pose and answer subtle questions and fully exploit the information contained in the images. Here we present iCluster, a complete integrated methodology for visualising and testing for difference in high throughput fluorescent protein subcellular localisation imaging. iCluster combines components for automated statistics generation, spatial organisation and layout of image sets by similarity, and a novel method of statistical testing to give rigorous confirmation of difference between experiments. Within iCluster, large image sets (up to 1400 image