SOFTWARE DVMT & DISSEMIN GEN OF MODELS FOR LARGE PROTEIN & NUCLEIC ACID SYSTEMS

此前，我们实验室开发的模型都是低 分辨率模型，每个模型用一个伪原子表示 16 S RNA的核苷酸和单个球形假原子 代表每一种蛋白质。 在过去的一年里，我们开始 核糖核蛋白的混合分辨率表示 粒子 第一个进步是引入了一种中间体， 分辨率蛋白质模型，每个氨基酸有一个假原子;这是 用于表示蛋白质晶体的适当的细节水平 核糖体的低分辨率模型中的结构，因为它 可以精确处理蛋白质的形状和方向。 第二 进步是引入了伪原子表示， 转移RNA（tRNA）和信使RNA（mRNA）。 前者被视为 作为一个刚性物体，其几何形状与tRNA晶体相同， 结构，而后者具有刚性的几何形状， mRNA/tRNA密码子/反密码子相互作用和外部灵活的几何形状 这六个核苷酸区域。 我们还开发了自动化 转换工具来提取适当的伪原子位置， RNA和蛋白质的全原子表示。 这些工具使我们 从晶体学、NMR或详细的 从全原子模型到中低分辨率模型 适用于YAMMP精炼。 他们是我们的重要组成部分， 努力开发NAB与 YAMMP模型。 利用刚刚描述的模型和工具的进步， 利用核糖体新数据的质量和数量， 我们正在与J. Frank合作开发下一代 核糖体的模型 我们评估了所有可用的数据， 从数据中建立了一组合适的约束条件，并生成 使用组合距离几何/ YAMMP的初步模型 细化方法。 我们希望完成这些第二代 在下一年的补助金。

Previously, the models developed in our laboratory were all low resolution models, with a single pseudoatom representing each nucleotide of the 16S RNA, and a single spherical pseudoatom representing each protein. During the past year, we have begun developing mixed resolution representations for ribonucleoprotein particles. The first advance was to introduce an intermediate resolution protein model, with one pseudoatom per amino acid; this is an appropriate level of detail for representing protein crystal structures in the low resolution models of the ribosome, because it allows accurate treatment of protein shape and orientation. A second advance was the introduction of pseudoatomic representations for transfer RNA (tRNA) and messenger RNA (mRNA). The former is treated as a rigid object whose geometry duplicates that of the tRNA crystal structure, while the latter has a rigid geometry in the region of mRNA/tRNA codon/anticodon interaction and a flexible geometry outside this six nucleotide region. We have also developed automated conversion tools to extract appropriate pseudoatom positions from all-atom representations of RNAs and proteins. These tools enable us to convert atomic models from crystallography, NMR, or detailed all-atom model building to low and intermediate resolution models suitable for YAMMP refinement. They are an essential component of our efforts at developing methods for interconversion between NAB and YAMMP models. Using the advances in models and tools just described, and exploiting the quality and quantity of new data of the ribosome, we are collaborating with J. Frank to develop the next generation models of the ribosome. We have evaluated all the available data, established a suitable set of constraints from the data, and generated preliminary models using the combined distance geometry / YAMMP refinement approach. We expect to complete these second generation efforts during the next year of the grant.