Theoretical Studies of Membrane Proteins

本项目的目的是开发和测试膜蛋白同源性建模的方法。可溶性蛋白和膜蛋白的同源性建模是非常独立发展的领域，部分原因是两者可用的结构信息数量存在巨大差异。在可溶性蛋白质的情况下，计算工具的发展和验证得益于大量三维结构的可用性。例如，可以建立模型，然后在CASP实验（结构预测的关键评估）的盲目预测中进行测试，或者以无偏的方式将已知结构的“预测”与实验结构进行比较。越来越多的膜蛋白结构变得可用，这表明对这类蛋白质进行类似的验证现在是可能的。这项研究的中心目标是开发一套新的软件工具，使研究人员能够建立模型，并以公正的方式对它们进行评估。该研究的一个关键要素是基于相关蛋白的坐标来构建已知结构的膜蛋白模型。将测试不同能量函数区分诱饵和天然构象的能力。预测蛋白质相对于脂质双分子层的正确方向的能力也将被测试。将要采用的建模方法是基于PI实验室在可溶性蛋白质研究中开发的方法。其中包括预测固定主链上氨基酸侧链构象的新工具，新的环预测方法和新的模型构建程序。该研究的一个关键要素是使用基于物理化学的能量标准来评估给定蛋白质的不同模型的相对构象能。随着研究的深入，将开发出适合膜蛋白的精细能量方法，并将其与来自不断增长的膜蛋白结构数据库的螺旋包装标准相结合。这项研究的更广泛影响包括软件工具的开发，这些工具可能被广泛用于基础研究和各种技术应用。这项研究还将为培训女性和少数族裔科学家提供一个平台，这是该实验室一段时间以来的传统。

The objective of this project is to develop and test methods for the homology modeling of membrane proteins. The homology modeling of soluble proteins and membrane proteins are areas that have developed quite independently, in part because of the vast difference in the quantity of structural information that has been available for both. In the case of soluble proteins the development and validation of computational tools have benefited from the availability of a large number of three-dimensional structures. For example, it has been possible to build models and then test them either in blind prediction as in the CASP experiment (Critical Assessment of Structure Prediction) or by comparing "predictions" of known structures to experimental structures in an unbiased fashion. The increasing number of membrane protein structures that are becoming available suggests that a similar kind of validation is now possible for this class of proteins. The central goal of this research is the development of a new battery of software tools that will allow researchers to build models and to evaluate them in an unbiased fashion. A key element of the research is the construction of models for membrane proteins of known structure based on the coordinates of related proteins. The ability of different energy functions to distinguish the decoy from the native conformation will be tested. The ability to predict the correct orientation of the protein with respect to the lipid bilayer will also be tested. The modeling approach to be taken is based on methods that have been developed in the PI's lab in studies of soluble proteins. These include novel tools to predict the conformation of amino acid side chains on afixed backbone, novel loop prediction methods and a new model building program. A key element of the research is the use of physical chemical-based energy criteria to evaluate the relative conformational energies of different models for a given protein. Refined energetic methods suitable for membrane proteins will be developed as the research progresses and will be combined with criteria for helix packing derived from the growing database of membrane protein structures. The broader impact of the research includes the development of software tools that are likely to be widely used in basic research and in a variety of technological applications. The research will also provide a platform for the training of women and minority scientists, a tradition in the lab for some time.