Turnkey solution for image phenotype classification

描述（由申请人提供）：荧光显微镜技术的最新进展使各种基于图像的分析得以发展，这些分析适用于药物发现、神经科学中的RNAi筛选、酵母遗传学和癌症研究等领域。这些检测富含表型信息，如细胞形态和蛋白质定位和活性。然而，尽管技术进步和丰富的蜂窝数据的新来源的承诺，重大的困难仍然在解释图像数据的研究人员。最值得注意的是，缺乏易于使用的工具来生成快速和全面的图像数据摘要。

DESCRIPTION (provided by applicant): Recent advances in fluorescence microscopy technology have enabled the development of a wide variety of image-based assays that are applicable to such areas as drug discovery, RNAi screens in neuroscience, yeast genetics, and cancer research. These assays are rich in phenotypic information such as cell morphology and protein localization and activity. However, despite the technological progress and promise of rich new sources of cellular data, significant difficulties remain in the interpretation of image data by researchers. Most notably, easy to use tools for producing rapid and comprehensive summaries of image data are lacking. A turnkey software package for the analysis of image data will be developed, tested, and deployed to the research community. The software will automatically classify experimental conditions by image phenotypes, and display specific examples of image phenotypes that distinguish experimental conditions. This classification will be able to group images either with or without prior knowledge of specific cell phenotypes. The software will be capable of running on a single desktop computer, requiring no specialized expertise for most biological applications. Thus, the goal of this proposal is to make the extraction of summarized information from large image data sets as routine as analyzing microarray data. The results of this work will provide a significant expansion of the capacity of fluorescence microscopy as a basic research tool, providing researchers with a new ability to inquire into such fundamental processes as the mechanisms of cellular signaling, and cellular responses to therapeutics such as sensitivity, resistance, and toxicity. Project narrative: In this proposal, a turnkey software tool for the automatic identification of important cellular phenotypes in images obtained by fluorescence microscopy will be developed, tested and delivered to the research community. This is important because general tools for producing rapid and compressive summaries of image data are lacking. The results of this work will provide a significant expansion of the capacity of fluorescence microscopy as a basic research tool, providing researchers with a new ability to inquire into such fundamental processes as the mechanisms of cellular signaling, and cellular responses to therapeutics such as sensitivity, resistance, and toxicity.