Multidimensional Quantitative Imaging Core

项目摘要/摘要 多维定量成像核心将提供定量图像、基因组和统计 对该中心研究人员产生的显微镜和测序项目进行分析。 以前，核心以计算生物学核心的形式运作，只专注于为 P50研究人员项目的基因组和统计分析。我们已经扩大了 核心将包括图像分析，因为P50调查人员的许多项目将检查细胞和 伴随分化、成熟和细胞谱系变化的亚细胞形态变化 规格。核心的目标之一是将深度学习中的新进展应用于图像识别 细胞结构。由于应用深度学习技术需要大量的计算资源 除了在图像采集和分析方面的专业知识外，Core设施更适合并配备更多 处理这些类型的分析比单独的调查人员或团队要好。Core正在与加州大学旧金山分校的 细胞构建中心将开发用于细胞图像分析的通用深度学习算法 生物学。核心还将通过协助开发De，将“组学”数据与成像数据相结合 Novo地图分析，以整合不同的数据集。最后，核心将参加培训和 教育P50学生、实习生和研究员生物信息学和图像分析的最佳实践。

PROJECT SUMMARY/ABSTRACT The Multidimensional Quantitative Imaging Core will provide quantitative image, genomic and statistical analysis for microscopy and sequencing projects generated by the Center's research investigators. Formerly, the Core operated as a Computational Biology Core and focused only on providing expertise for genomic and statistical analysis for the P50 investigators' projects. We have expanded the scope of the Core to include image analysis since many of the P50 investigators' projects will examine cellular and subcellular morphological changes that accompany changes in differentiation, maturation and cell lineage specification. One of the goals of the core is to apply new advances in deep learning to image recognition of cellular structures. Since applying deep learning techniques requires significant computational resources as well as expertise in image acquisition and analysis, the Core facility is better suited and equipped to handle these types of analyses than individual investigators or teams. The Core is partnering with UCSF's Center of Cellular Construction to develop general deep learning algorithms for image analysis of cell biology. The Core will also integrate “omics” data with imaging data by aiding in the development of de novo mapping analyses to integrate the disparate datasets. Finally, the Core will participate in training and educating P50 students, interns and fellows in the best practices of bioinformatics and image analysis.