Development of an accessible integrated computational protein design software suite

本研究计划的目标是建立在我们现有的方法，即正常模式分析方法ENCoM和FlexAID (Gaudreault et al., 2014)，开发一个集成的计算蛋白设计（CPD）软件套件，与现有软件竞争，但不同于现有的CPD软件，易于使用；让非专业人士也能接触到计算蛋白质设计。我们之前表明，基于熵的ENCoM方法非常适合于稳定突变的预测（Frappier et al., 2014 & 2015）。这导致我们在与Amai蛋白的现有合作中利用它来帮助他们提高甜蛋白的热稳定性。本程序的第一个目标是在一个集成软件中引入不同的焓预测方法。与CPD相关的一个重要问题是，当考虑多个突变位点上的多个侧链构象以及主链灵活性时，搜索空间的巨大尺寸。现有的方法仅限于非常少量的突变和最小的骨干运动，因为它们的目标是达到可证明的全局最小能量配置（GMEC）。我们的第二个目标是利用已经用于对接软件FlexAID的遗传算法（GA）搜索方法来执行概率搜索，以允许更广泛的序列和构象空间搜索。第三个目标是在遗传适应度函数中整合溶解度的预测和消除二硫键的过程，同时保持稳定性和灵活性。我们的第四个目标是将上述方法集成到用户友好的NRGsuite GUI中（Gaudreault, 2015）。我们将利用人乳铁蛋白作为实验模型系统。乳铁蛋白在人类健康和食品技术方面有着广泛的应用，其中包括其广泛的抗菌特性（Brock, 2011）。Amai将对我们的预测进行进一步的实验验证，以优化thumatin和monellin。在提出的新型CPD套件中，使用ENCoM作为中心工具的一个独特优势是，它允许我们测量突变对蛋白质结构所有部分的动态影响。这反过来又让我们引入可能改变稳定性的突变，而不影响蛋白质特定区域的动态，这些区域可能对特定功能很重要。上述项目在使用我们现有的工具方面具有高度的创新性，并将产生一个高性能的计算蛋白质设计软件套件，可供所有人使用。最后，开发更稳定形式的乳铁蛋白和甜蛋白的潜在好处除了在实验上验证我们的计算预测外，还可能预示着在人类健康和食品技术方面的重要应用。这些实验应用程序还将指导我们的软件功能的进一步发展，以满足学术和工业用户的需求。

The goal of this research program is to build upon our existing methods, namely the Normal Mode Analysis method ENCoM and FlexAID (Gaudreault et al., 2014), to develop an integrated computational protein design (CPD) software suite that performs competitively with existing software but unlike existing CPD software, is easy to use; Making computational protein design accessible to non-experts. We previously showed that the entropy-based ENCoM method is well suited for the prediction of stabilizing mutations (Frappier et al., 2014 & 2015). This led to its utilisation in our existing collaboration with Amai proteins to help them increase the thermal stability of sweet proteins. The first goal of this program is to introduce different enthalpy prediction methods within an integrated software. One important issue related to CPD is the massive size of search space when considering multiple side-chain conformations over several mutation sites as well as backbone flexibility. Existing methods are limited to a very small number of mutations and minimal backbone movements as they aim at reaching provable global minimum energy configurations (GMEC). Our second goal is to utilize our genetic-algorithm (GA) search method already used for our docking software FlexAID to perform probabilistic searches to allow a  wider search of sequence and conformational space. The third objective is to integrate within the GA fitness function the prediction of solubility and a procedure to eliminate dissulfide bonds while maintaining stability and flexibility. Our fourth objective is to integrate the above methods within our user-friendly NRGsuite GUI(Gaudreault, 2015). We will utilize human lactoferrin as experimental model system. Lactoferrin has ample applications in human health and food technology due to among others, its broad antibacterial properties (Brock, 2011). Additional experimental validation for our predictions will be performed by Amai on the optimisation of thaumatin and monellin. One unique advantage of the proposed use of ENCoM as a central tool within the proposed novel CPD suite is that it allows us to measure the effect of mutations on dynamics across all sections of a protein structure. This in turn lets us to introduce mutations that may change stability without affecting the dynamics in specific areas of the protein that may be important for a particular function. The above project is highly innovative in its use of our existing tools and will produce a high-performing computational protein design software suite accessible to all. Lastly, the potential benefits from developing more stable forms of lactoferrin as well as sweet proteins may herald important applications in human health and food technology in addition to serving to validate experimentally our computational predictions. These experimental applications will also guide further development of the capabilities of our software to suit the needs of both academic and industrial users.