| Image Informatics and Computational Biology Unit |
| Ilya G. Goldberg, Ph.D., Chief |
| Overview: Our goal is to develop quantitative visual assays for functional genomics. Visual assays (fluorescence microscopy, for example) represent the most general way we can assess gene function, and thus they have been at the core of experimental biology for the past 400 years. The main limitation in applying visual assays to functional genomics is that they are laborious, and often qualitative, subjective or anecdotal. Currently, our work is split into two areas: Development of computational tools and information systems for functional genomics, and the application of these tools to two specific biological areas: Determining the spatial distribution of differentially expressed gene products in pre-implantation mouse embryos, and visual screening of RNAi libraries in C. elegans. |
| In collaboration with Minoru Ko's group (Developmental Genomics and Aging Section), we are using restoration microscopy to image whole-mount mouse embryos processed for in situ hybridization with probes for selected differentially expressed genes. The goal of this project is to determine the earliest molecular events leading to asymmetry and differentiation in the pre-implantation mouse embryo. A longer-term goal of this work is to integrate various experimental data sources about mouse genes with existing sources of mouse genomic information such as sequences and their annotations. In order to accomplish this goal, we are developing information systems that can represent the full range of available experimental data. Once these experiments can be formally described and made computationally accessible, data from very different kinds of experiments can be normalized and compared against one another through their common elements - the genes and proteins the experiments are performed on. |
| Our goal with visual screening of RNAi libraries is to establish quantitative image analysis protocols that will allow rapid turnover of visual genetic screens. Our aim is to automate the screening and analysis effort to such an extent that a full-genome visual RNAi screen becomes a routine experiment. Once such screens become routine, it will be possible to cross-reference them so that they act as positive and negative controls for one another. This 'sea of screens' can then be used for in silico experiments and for de novo reconstruction of genetic interaction pathways. |
| We are also principal developers and co-founders of the Open Microscopy Environment (OME) project. OME is a software package and a set of standards for image informatics - the collection, maintenance, and analysis of biological images and associated data. The aim of this project is to standardize how image information is stored, extracted and transported between different software applications. At the core of OME is a database that maintains images and all associated image meta-data including experimental descriptions and image processing results. OME also specifies XML-based data interchange standards that allow software packages from different microscope manufacturers to exchange not only images, but also experimental descriptions, image acquisition details and results of any image processing. Finally, OME provides an extensible framework for modular image analysis that allows rapid development of image analysis work-flows. OME is intended to address a wide variety of applications from high-resolution imaging used for detailed mechanistic studies to image-based screening applications involving hundreds of thousand of images. |
| Staff: |
|
John Delaney, Ph.D.
Nikita Orlov, Ph.D.
Lior Shamir
David Mark Eckley, Ph.D.
Christopher Coletta
Salim Rahimi
|
|
|
