ABI Innovation: Robotics-inspired modeling and design of proteins

加州大学旧金山分校（University of California San Francisco）获得了一笔资助，用于研究捕获和设计蛋白质结构变化的方法。这对于使设计更精确和设计需要结构变化的功能都很重要。计算蛋白质设计具有巨大的潜力，可以通过创造具有有用功能的蛋白质来改变科学和工程的许多领域：构建可以产生重要化学物质和燃料的生物合成途径的新酶；处理生物信息的选择性信号分子；新的蛋白质材料具有复杂的功能，可以模仿它们的生物对应物。这项研究旨在开发计算方法，最终使这种复杂的蛋白质功能可设计。具体而言，该项目将产生(i)模拟酶特异性变化和改进活性位点设计的方法；（ii）建模和设计作为开关的信号蛋白的方法；（iii）生物学家和生物工程师可用的基准和文件协议。最近的设计成功创造了新的酶、蛋白质-蛋白质相互作用和纳米材料，这激发了学术界和工业界对具有改良的、新的和有用功能的蛋白质的兴趣。结合技术能力的相当大的扩展（例如大规模的从头构建和生物“部分”的表征），将会越来越需要更多的预测设计方法，例如这里开发的方法，以创造具有新的和更复杂活动的蛋白质。应用范围广泛，包括计算和结构生物学，代谢工程，合成和细胞生物学。在这项资助下开发的新方法将用于分子工程的研究生课程和团队项目，旨在促进生物和物理/工程科学学生之间的合作，以及UCSF的本科和高中研究活动。改进的计算方法将通过Rosetta计算工具套件（www.rosettacommons.org）和易于访问的web应用程序作为源代码广泛传播。

The University of California San Francisco is awarded grant advance approaches that can capture and engineer structural changes in proteins. This is important both to make designs more accurate and to engineer functions that require structural changes. Computational protein design has enormous potential to transform many areas of science and engineering by creating proteins with useful functions: new enzymes to build biological synthesis pathways that can produce important chemicals and fuels; selective signaling molecules that can process biological information; and new protein materials with sophisticated functions mimicking their biological counterparts. This research seeks to develop computational methods to ultimately make such complex protein functions designable. Specifically, project will produce (i) methods to model changes in enzymatic specificity and improve the design of active sites; (ii) methods to model and design signaling proteins that act as switches; and (iii) benchmarked and documented protocols available to biologists and biological engineers.Recent design successes that created novel enzymes, protein-protein interactions, and nanomaterials are spurring both academic and industrial interest in proteins with modified, new, and useful functions. Combined with a considerable expansion of technological capabilities (such as large-scale de novo construction and characterization of biological "parts"), there will be an increasing need for more predictive design methods, such as the ones developed here, to create proteins with new and more complex activities. Applications span broad areas including computational and structural biology, metabolic engineering, and synthetic and cellular biology. The new methods developed under this grant will be used in graduate courses and team projects in molecular engineering designed to foster collaboration between students from the biological and physical/engineering sciences, as well as in undergraduate and high-school research activities at UCSF. Improved computational methods will be disseminated broadly as source code via the Rosetta suite of computational tools (www.rosettacommons.org) and via easily accessible web applications.