Theoretical Condensed Matter Physics

这笔赠款由材料研究和分子与细胞生物科学部联合支持。所有多细胞生物都是从未分化的单细胞卵子开始的，并利用基因组中的指令来创造成年动物的特化细胞和器官。最近的基因组测序项目强化了早期的观点，即形成源于基因是如何调控和协调的；由此可以看出，进化更多地是通过对基因的调控进行修修补补，而不是创造新的基因。从基因组推断基因调控仍处于初级阶段，目前的许多基因组测序项目都集中在相关生物的簇上，目的是使用物种间比较作为筛选哪些是保守的，也就是说对功能最重要的。然而，我们在很大程度上仍然不知道调控信息在基因组中是如何编码的，简单地比较序列可能就像用两种语言比较同一个句子一样。统计力学基本上关注的是计算模式的概率和评估样本是否与模型一致(更专业地说，是给出模型的数据的概率)。100年来，果蝇一直是遗传和发育的模式，10个相关物种的基因组将于2005年初公布。我们将与洛克菲勒大学的一个发育生物学实验室一起，询问我们在计算基因组哪些区域控制早期胚胎模式方面的早期成功，是否可以通过计算将调节蛋白与DNA结合的分配函数来使其更加定量。我们在苍蝇中了解的调控将被映射到其他测序物种以及蚊子上，作为对我们理解的测试，并看看调控是如何演变的。在之前的一个项目中，我们通过研究最近分化的苍蝇物种，对调控序列发生变化的基本分子事件进行了分类。由于调控DNA通过结合蛋白质来指导基因表达，模拟这种特性在突变和选择下是如何保持的，应该会为如何破译自然序列提供线索。其目标是将进化塑造为一个优化过程，从而使其更多地成为一种预测性理论，而不是历史理论。细胞必须协调许多过程才能分裂，细胞必须合作才能形成有机体。完成这种协调的网络是当前生物学的一个重要焦点。细胞分裂周期可以说是生命本身的基础，许多关键基因在单细胞酵母和哺乳动物之间保存着。细胞分裂和生长控制中的异常会导致癌症。PI计划用定制的图像处理软件处理单个酵母细胞生长成簇的电影，以研究细胞到细胞的变异性如何指示控制网络。更大的细胞是否分裂得更快，细胞周期不同阶段的持续时间是否相关？细胞周期必须严格遵守某些过程的顺序，例如DNA复制必须先于染色体的分离，我们将询问这种波动是否为细胞周期中的子模块提供了证据(例如，DNA复制是否与芽的出现有关，以及某些周期蛋白基因的表达)。大量的细胞周期突变已经从对细胞群体的筛选中衍生出来，人们应该在单个细胞水平上重新检查它们的性质。具体地说，是否某些基因负责维持子模块的时间一致性？作为一个具有代表性的蜂窝网络，细胞周期的某些方面是否可以更好地模拟为离散的、二进制的系统，而不是一组微分方程式？更广泛的影响：增强研究和教育的基础设施。PI在纽约市的洛克菲勒大学和纽约伊萨卡的康奈尔大学之间奔波，并与许多物理科学和生物学方面的个人合作。他是许多生物学学生的论文委员会成员，也是NIH资助的数量生物学中心的顾问小组成员，以及为过渡性研究提供资助的基金会的顾问小组成员，在那里，他通常是唯一的量化成员。这项提议将资助想要进入生物专业的自然科学专业的学生。我们通过定制的网站提供我们的计算预测，我们的软件包也可以免费分发。

This grant is supported jointly by the Divisions of Materials Research and Molecular and Cellular Biosciences. All multi-celled organisms begin as an undifferentiated single celled egg and use the instructions written in the genome to create the specialized cells and organs of the adult. Recent genome sequencing projects reinforce earlier ideas that form is derived from how genes are regulated and coordinated; and by implication evolution, proceeds more by tinkering with the regulation of genes than creating new genes. The inference of gene regulation from the genome is still in its infancy, and many of the current genome sequencing projects are focused on clusters of related organisms with the intent of using interspecies comparisons as a filter for what is conserved and by implication most important for function. However, we are still largely ignorant of how regulatory information is encoded in the genome and simply comparing sequence may be like comparing the same sentence in two languages. Statistical mechanics is fundamentally concerned with computing the probabilities of patterns and assessing whether a sample is consistent with a model (more technically, the probability of the data giventhe model). The fruit fly has been a model for genetics and development for 100 years, and the genomes of 10 related species will be available in early 2005. Together with a developmental biology lab at Rockefeller University, we will ask whether our earlier successes in computing which regions of the genome control early embryonic patterning, can be made more quantitative by computing the partition function for binding the regulatory proteins to the DNA. The regulation we understand in fly will be mapped onto the other sequenced species, as well as the mosquito, as a test of our understanding and to see how regulation evolves.In a previous project we classified the basic molecular events through which regulatory sequence changes by studying recently diverged fly species. Since regulatory DNA directs gene expression by binding proteins, simulating how this property is preserved under mutation and selection should provide clues about how to decode natural sequence. The goal is to cast evolution as an optimization process and thus make it more a predictive theory, rather than a historical one.Cells have to coordinate many processes to divide and cells must cooperate to form an organism. The networks that accomplish this coordination are an important focus of current biology. The cell division cycle is arguable the basis of life itself, and many of the key genes are preserved between single celled yeast and mammals. Aberrations in the control of cell division and growth lead to cancer.The PI plans to process movies of single yeast cells growing into clusters with custom image processing software to investigate how cell to cell variability is indicative of the control network. Do larger cells divide more rapidly, and does the duration of the various phases of the cell cycle correlate? The cell cycle has to strictly order certain processes, e.g. DNA replication must precede the segregation of chromosomes, we will ask whether the fluctuations provide evidence for sub modules within the cell cycle (e.g. is DNA replication tied to bud emergence, and the expression of certain cyclin genes). A large collection of cell cycle mutants has been derived from screens on populations of cells and one should reexamine their properties at the single cell level. Specifically are certain genes responsible for maintaining the temporal coherence of submodules? As a representative cellular network, are there some aspects of the cell cycle that are better modeled as discrete, binary, systems rather than sets of differential equations?Broader Impacts: Enhance Infrastructure for Research and Education.The PI divides his time between The Rockefeller University, in New York City, and Cornell University, in Ithaca, NY, and collaborates with many individuals in the physical sciences and biology. He is on the thesis committee of many biology students, and the advisory panels for NIH funded centers in quantitative biology and foundations awarding grants for transitional research, where he is often the only quantitative member. This proposal will fund physical science students wanting to move into biology. We make our computational predictions available through customized web sites, and our software packages are distributed freely.