skip to primary navigationskip to content
 

High-Content Microscopy and Big Data: Discovering the genes and pathways that control cells in health & disease

last modified May 20, 2015 05:43 PM
Rafael Carazo Salas, Department of Genetics

High-­‐Content Microscopy and Big Data: Discovering the genes and pathways that control cells in health & disease

Rafael E Carazo Salas, Genetics Department (contact: cre20@cam.ac.uk)

Other authors: Veronika Graml, Xenia Studera, Jonathan Lawson, Anatole Chessel, James Dodgson, Federico Vaggi, Marco Geymonat, Laura Wagstaff, Bálint Antal, Marco Giordan, Miriam Bortfeld-Miller, Miki Yamamoto, Kunio Arai, Juan F Abenza, Thomas Walter, Masamitsu Sato, Attila Csikasz-Nagy, Eugenia Piddini

Abstract

Understanding how complex phenotypic traits arise from the genome is both the promise and challenge of modern biology. It has the potential to identify the origin of many diseases and inspire Personalised Medicine strategies to combat them. A necessary step towards that understanding is being able to extract and integrate detailed phenotypic information at the cellular level, such that it can be interpreted unequivocally and quantitatively with respect to genotypic information.

Automated high-throughput microscopy-based technologies provide an increasingly important tool to understand how genotype engenders phenotype. In the past 5 years our group has pioneered the development of 3D image-based high-throughput/high-content (HT/HCM) microscopy phenotyping pipelines for functional genomics, using yeast as experimental model. Capitalizing on this technology, we have carried out diverse screens for genes and pathways regulating key cell biological processes (Graml et al., Developmental Cell 2014, and http://www.sysgro.org/; Jeffares et al., Nature Genetics 2015; Geymonat et al., under review, Developmental Cell; Antal et al, submitted; Dodgson et al, in preparation), discovering and annotating hundreds of new genes and many functional links between processes.

For example, we carried out a genome-wide screen that discovered 100s of mostly conserved genes involved in controlling and linking cell morphology, the microtubule cytoskeleton, cell cycle progression and other processes, including a novel link between the DNA damage response and microtubule control (Graml et al., Developmental Cell 2014, and http://www.sysgro.org/). As drugs targeting microtubules and DNA damage are highly relevant in chemotherapy, a mechanistic dissection of this novel link could be of biomedical relevance. In another project, we used this type of approach to wire the network that controls cellular polarity, a basic process of deep biomedical relevance (Geymonat et al., under review, Developmental Cell; Dodgson et al, in preparation). Our vision with these studies is to generate an unprecedented and incremental genome annotation resource, providing fundamental new insights into how genes regulate multiple biological processes and how processes are co-regulated.

In order to make the Big Data generated from HT/HCM cellular phenotyping projects accessible and further mineable by the community, and to potentiate the return-on-investment of those projects, we also have began developing web-based visual analytics tools allowing noncomputational scientists to access and mine the large (>106 points) datasets we are producing (Antal B et al, submitted; http://vizbi.org/Posters/2014/E02/; http://www.mineotaur.org/).

Here, I present the development of our microscopy phenomics pipeline, selected results from the two example projects mentioned and our biological Big Data vision.