Determining the Cyberinfrastructure Needs for Efficient Phenomics Research

威斯康星大学获得了一笔赠款，用于为一个研究所进行规划，通过开发软件和高通量计算方法来捕获和分析生物形式和功能的数据，以推进表型组学研究。生物学研究的一个主要目标是了解一个生物体--S的互补基因(基因组)如何对该生物体(表型)的出现、生长、发育、行为等作出贡献。许多表型信息可以在生物生长和发育的数字图像中捕捉到，这些生物的规模从微观细胞到作物田的整个植物。人们迅速认识到图像分析可以如何有效地推进基因型与表型关系的研究，这增加了对改进计算方法的需求的紧迫性。目前，采用这种实验方法的研究实验室都在创建自己的定制网络基础设施，用于收集和分析图像，以实现必要程度的自动化和吞吐量。这个概念化项目将把生物学家和计算机科学家聚集在一起，研究促进基于图像的表型研究所需的网络基础设施(CI)的发展。要探索的想法可以归结为一般性问题。如果没有计算机科学人员的持续支持，如何才能使表型研究的计算方面足够例行公事，以便生物研究小组使用？不同的研究环境(例如，主要是本科生或工业园区)是否需要不同的解决方案？如何才能增加生物学家自然倾向于并在学术上准备好在表型研究所在的计算/数学/生物学界面工作的研究人员的数量？将在关键会议上举办应邀参加的讲习班和公开活动，以进行知情的辩论。一个由具有不同经验的生物学家和计算机科学家组成的指导委员会将协调这些活动，并将调查结果综合成一份公开报告，其中将包括关于如何创建有效和易于使用的网络基础设施的建议，以促进这一重要的生物学研究领域。植物生物学研究通常是一个例子，但任何生成和分析大型高维数据集的领域都将从这个项目中受益，其中层叠和层层的数字图像只是一个例子。使该项目产生广泛影响的另一个方面是对下一代生物学研究人员的教育和准备的重视。一般来说，生物专业的学生必须为生物/计算/数学界面的工作做好更好的准备，否则即使是最好的CI也只会被受益于它的研究界的一小部分人利用。

The University of Wisconsin is awarded a grant to conduct planning for an institute to advance research in phenomics through the development of software and high-throughput computational methods for capturing and analyzing data on biological form and function. A major goal in biology research is to understand how an organism?s complement of genes (the genome) contributes to the appearance, growth, development, behavior, etc. of the organism (the phenotype). Much phenotype information can be captured in digital images of organisms undergoing growth and development at scales ranging from microscopic cells to whole plants in crop fields. The rapidly growing realization of how effectively image analysis can advance the study of genotype-to-phenotype relationships adds some urgency to the need for improved computational methods. Currently, research labs adopting this experimental approach are each creating their own custom cyber infrastructures for collecting and analyzing images to achieve the necessary degree of automation and throughput. This conceptualization project will bring biologists and computer scientists together to investigate the cyber infrastructure (CI) developments needed to facilitate image-based phenotype research. The ideas to explore can be cast into general questions. How can the computational aspects of phenotype research be made routine enough for a biology research group to use without continuous support of a computer science staff? Do different research contexts (e.g. primarily undergraduate or industrial campuses) require different solutions? What can be done to increase the number of research biologists naturally inclined and academically prepared to work at the computation/math/biology interface where phenotype research resides? Invited-participation workshops and open events at key conferences will be conducted to generate informed debate. A Steering Committee consisting of diversely experienced biologists and computer scientists headed by the PI and assisted by staff will orchestrate the activities and synthesize the findings into a public report that will include recommendations about how to create an effective and easy-to-use cyberinfrastructure that would advance this important area of biology research. Plant biology research will often be the example but any field that generates and analyzes large, high-dimensional datasets, of which stacks and layers of digital images are just one example, stands to benefit from this project. Another aspect that gives this project a broad impact is the emphasis on education and preparation of the next generation of biology researchers. Biology students in general must become better prepared for work at the biology/computation/math interface or even the very best CI will be utilized by only a narrow portion of the research community that stands to benefit from it.