Dublin, Ireland (PressExposure) June 20, 2014 -- Robert Murphy, Carnegie Mellon University, will give a keynote address at the Cell Based Assay & Screening Technologies Conference, to be held October 7-8, 2014 in Dublin, Ireland.
Dr. Murphy, who currently serves as a professor and the director of the Ray and Stephanie Lane Center for Computational Biology in the School of Computer Science at Carnegie Mellon University, focuses his research on combining fluorescence-based cell measurement methods with quantitative and computational methods. His group at Carnegie Mellon has pioneered a number of machine learning methods for analyzing and modeling various aspects of cell organization.
Traditional HCS methods detect perturbed phenotypes but do not identify the specific changes in cell organization underlying them; thus, results also cannot be easily compared across different HCS systems. Dr. Murphy will share about his group's work to develop an open source system that is able to convert images into generative models of cell components that capture their properties and how they relate to each other in a given cell type. These generative models are capable of producing new instances of a pattern that are statistically similar to those it was learned from; they thus capture the biological reality behind the images as well as cell heterogeneity. Furthermore, because parameters of the model are highly interpretable and independent of the microscope system, they are ideal for analyzing perturbations in high content screening applications.
Once a common framework for representing the effects of perturbagens can be created, the next step is to learn a comprehensive model that describes the effect of large numbers of potential perturbagens (e.g., small molecule compounds). Though Dr. Murphy and his group could try to build such a model by measuring all combinations of perturbagens and targets, the experimentation required would be enormous. Dr. Murphy will therefore share about how his group has instead developed "active" machine learning approaches that iteratively select experiments to perform in order to improve the best model currently available. Results in test cases show that very accurate models can be built with significantly fewer measurements than exhaustive screening.
For more information on Dr. Murphy's presentation, as well as the conference, please visit http://www.gtcbio.com/cbast2014.