Flexible Macromolecular Docking

蛋白质的完全柔性的肽、环或末端与相对结构化的位点的结合 在其分子伴侣的表面上是一种普遍存在的瞬时大分子 介导大多数识别和信号级联的相互作用。当跨越一个 广泛的亲和力，这种相互作用统一涉及结构化和非结构化 合伙人，在彼此不存在的情况下可能作为独立实体存在， 在有限的时间内，导致构象变化， 其它蛋白质进入/离开复合物，以及下游途径的激活或抑制。 蛋白质-肽相互作用的合成肽调节剂通常是有前途的 癌症、炎症和内分泌失调的治疗候选物。 尽管瞬时蛋白质-肽相互作用对于生物医学研究至关重要 和治疗发现，只有一小部分这些复合物是适合的， 实验结构测定因此，只有精确的肽对接才可能导致 在蛋白质-肽相互作用的结构理解方面取得了突破。然而对于更长的肽 超过6-8个氨基酸，构象搜索的压倒性大小和复杂性 空间，不可避免的不准确的结合位点表示，由于诱导拟合，效率低下 或不充分彻底的采样，以及评分函数错误的积累， 通过计算方法精确确定肽结合位姿和相互作用。 本提案旨在显著扩大肽和蛋白质大小的范围，以用于制备肽和蛋白质。 该精确的复杂几何形状预测可以通过全局构象 优化.这一进展将通过追求两个具体目标来实现：（目标1）发展 和优化的可靠的肽交叉对接程序使用化学场增强 随机整体中的结合位点表示和改进的熵计算 内部坐标的构象搜索;和（目标2）扩展开发的协议 具体的生物项目与实验验证的预测几何形状， 肽变体。目标1中提出的创新战略包括使用新的力场， 丰富的结合位点表示与化学领域，和最佳的 用于诱导配合的袋的构象膨胀。目标2的目标包括复合物 A类和B类GPCR及其蛋白质和肽调节剂以及G蛋白αi相互作用 与GEF肽。 实现这一建议的目标将导致在灵活性方面取得突破性进展。 大分子对接它将产生有价值的软件工具、协议和共享资源 对于生物界来说。这也将导致发现新的肽调节剂， 重要的治疗、免疫和诊断靶点。

The binding of a fully flexible peptide, loop, or terminus of a protein to a relatively structured site on the surface of its molecular partner is a prevalent type of a transient macromolecular interaction that mediates the majority of recognition and signaling cascades. While spanning a wide range of affinities, such interactions uniformly involve a structured and an unstructured partner that may exist as independent entities in the absence of one another and only associate for a limited period of time resulting in conformational changes, recruitment or dissociation of other proteins to/from the complex, and activation or inhibition of the downstream pathways. Synthetic peptide modulators of protein-peptide interactions are frequently promising therapeutic candidates in cancer, inflammation and endocrine disorders. Despite the critical importance of transient protein-peptide interactions for biomedical research and therapeutic discovery, only a small fraction of these complexes are amenable to experimental structure determination. Therefore, only accurate peptide docking may lead to a breakthrough in structural understanding of protein-peptide interactions. Yet for peptides longer than 6-8 amino acids, the overwhelming size and complexity of the conformational search space, the inevitable inaccuracies in the binding site representation due to induced fit, inefficient or insufficiently thorough sampling, and accumulation of scoring function errors prohibit the accurate determination of peptide binding poses and interactions by computational methods. The present proposal intends to dramatically expand the range of peptide and protein sizes for which accurate complex geometry prediction can be achieved by global conformational optimization. This advance will be made by pursuing two Specific Aims: (Aim 1) Development and optimization of a reliable peptide cross-docking procedure using chemical field-enhanced binding site representations and improved entropy calculations in the stochastic global conformational search in internal coordinates; and (Aim 2) Extension of the developed protocol to specific biological projects with experimental validation of the predicted geometries and peptide variants. The innovative strategies proposed in Aim 1 include using a new force field, the enrichment of binding site representations with chemical fields, and the optimal conformational expansion of the pocket for induced fit. The targets of Aim 2 include complexes of Class A and B GPCRs with their protein and peptide modulators and G-protein αi interactions with GEF peptides. The attainment of the aims of this proposal will lead to breakthrough advances in flexible macromolecular docking. It will result in valuable software tools, protocols and shared resources for the biological community. It will also lead to discovery of new peptide modulators of important therapeutic, immune, and diagnostic targets.