Mechanisms of Protein Structure Evolution

描述(申请人提供)：蛋白质进化的传统观点认为，所有的蛋白质结构域都是不同进化线的后代，这样的线相对较少(约1000条)，这些线都是相对古老的起源。两个新的证据使这一观点站不住脚。首先，对一组全序列基因组的分析表明，大多数蛋白质家族似乎都很小，而且分布在系统发育空间，显然意味着最近出现了。其次，对已知蛋白质结构之间的关系的分析表明，有1000多个不同的折叠，似乎意味着更多的进化路线。理论和事实之间的巨大差异已经清楚地存在了一段时间，并提出了许多解释。但到目前为止，还没有关于替代方案的明确研究。在这个项目中，我们将系统和定量地研究四个独立的假说，每个假说都可能占大量蛋白质折叠和明显年轻蛋白质的一定份额：这些假说(1)是新的开放阅读框架或框架移位的旧蛋白质的结果；(2)是旧蛋白质部分广泛重组的结果；(3)是从系统发育空间的未探索部分横向转移的；(4)是较大的较老家族的一部分，由于序列变化迅速，因此并不是所有的亲属都被发现。为了研究这些假设，我们将开发和扩展一套计算方法。这些方法包括构建蛋白质家族的方法；可靠地估计蛋白质家族的年龄，检测横向基因转移效应；确定用序列方法可能在多大程度上检测到家族成员；更定量地确定蛋白质结构是否在进化上相关；搜索远程结构和序列关系；以及分析作为家族年龄函数的一系列蛋白质性质。我们还将建立一个网络资源，用于发布结果并征求广泛的社区注释和讨论。 与公共卫生相关：了解蛋白质分子的结构和功能适应性特性是医学的许多方面的基础，特别是新病毒的出现、药物设计和对抗传染病新疗法的耐药性。

DESCRIPTION (provided by applicant): The traditional view of protein evolution has been that all protein domains are descendents of distinct evolutionary lines, that there are a relatively small number of such lines (about 1000), and that these lines are all of relatively ancient origin. Two new bodies of evidence make that view untenable. First, analysis of sets of fully sequenced genomes shows that most protein families appear to be small, and narrowly distributed in phylogenetic space, apparently implying recent emergence. Second, analysis of the relationship between known protein structures shows that there are many more than a 1000 distinct folds, appearing to imply many more evolutionary lines. The large discrepancy between theory and fact has been clear for some time and a number of explanations have been put forward. But so far there has been no definitive study of alternatives. In this project we will systematically and quantitatively investigate four separate hypotheses, each of which may account for some share of the large number of protein folds and apparently young proteins: that these (1) are the result of generation of new open reading frames or frame shifted older ones; (2) are the outcome of extensive recombination of portions of older proteins; (3) are laterally transferred from unexplored parts of phylogenetic space; (4) are part of larger older families where there has been rapid sequence change, such that not all relatives are found. To investigate these hypotheses we will develop and extend a set of computational methods. These include methods of building protein families; reliably estimating the age of protein families, detecting lateral gene transfer effects; determining to what extent members of families are likely to have been detected with sequence methods; more quantitatively determining whether protein structures are evolutionarily related; searching for remote structure and sequence relationships; and analyzing a range of protein properties as a function of family age. We will also construct a web resource for distributing the results and soliciting extensive community annotation and discussion. PUBLIC HEALTH RELEVANCE: Understanding the structural and functional adaptive properties of protein molecules underpins many aspects of medicine, particularly the emergence of new viruses, drug design, and combating resistance to new therapeutics in infectious diseases.