CAREER: Global Self-Organization of all Known Proteins - Toward a Complete Map of the Protein Space

EIA-0133311戈兰约纳康奈尔大学CAREER：所有已知蛋白质的全球自组织--走向蛋白质空间的完整图谱这项努力满足了对所有已知蛋白质(蛋白质空间)进行表征和组织成功能类别的日益增长的需求。这项工作将开发敏感的算法来检测蛋白质和蛋白质家族之间的微妙关系，并将研究和开发将蛋白质空间表示为可以进一步探索的数学对象的新技术。此外，它还将开发新的无监督学习技术，用于研究蛋白质空间的性质和几何结构，并最终创建完整的蛋白质空间路线图。MAP的一个直接好处是能够自动对每一种蛋白质进行分类；已知的蛋白质以及未来将被测序的新蛋白质。这种映射可以根据序列在空间中的相对位置来指示新确定的序列的可能功能，而这些新确定的序列是传统方法无法破译的。它可以揭示远亲蛋白质家族之间的新关系，限制序列可能的结构构象，解释序列到结构的映射，并可能增加对基本进化过程的洞察。

EIA-0133311Golan YonaCornell UniversityCAREER: Global Self-Organization of all Known Proteins - Toward a Complete Map of the Protein SpaceThis effort addresses the growing need to characterize and organize all known proteins (the protein space) into functional classes. This effort will develop sensitive algorithms to detect subtle relationships between proteins and protein families, and will study and develop new techniques for representing the protein space as a mathematical object that can be further explored. Additionally, it will develop new unsupervised learning techniques for studying the properties and geometry of the protein space, and eventually to create a complete road map of the protein space.The outcome proposed will be a complete, accurate map of the protein universe. A direct benefit of the map is the ability to automatically classify every protein; known proteins as well as new proteins that will be sequenced in the future. This map can indicate the possible function of newly determined sequences that cannot be deciphered by traditional methods, based on the relative position of the sequence in the space. It can expose new relationships between distantly related protein families, limit the possible structural conformations of sequences and explain the mapping of sequences to structures, and may add insights about fundamental evolutionary processes.