计算设计以信封蛋白为靶点的抗登革热病毒肽抑制剂与初步实验验证

Dengue fever is mosquito-borne disease caused by Dengue Virus（DENV）. There are about 390 million infection cases of dengue fever each year,there is no effective vaccine for DENV. The infection by DENV is initiated by fusion of the viral and cellular membranes, the self-inhibitory peptides derived from the envelope(E) protein function to inhibit the protein-protein interactions, thus they may have the potential to be developed into effective antiviral therapy. This project will focus on designing of anti DENV peptide inhibitors from E protein.To begin with, based on the Monte-Carlo computational method, the statistical potentials added with E proteins could show paramount accuracy compared with non-specific statistical potentials.Moreover, Evolutionary algorithm is effective to further optimize peptides obtained from the Monte-Carlo algorithm. Futhermore,we developed support vector machine models which elucidated the relationship between feature properties of E peptide inhibitors and antivirus activity,and using the models to screen the potential peptide inhibitors could increase the pertinence.Finally,biology experiment would be used for the validation，which mainly include treatment of cells with peptide inhibitors for pre or post-infection of DENV and toxicity of cells with peptide inhibitors. Therefore, those peptide inhibitors could be developed as a leading compound for the treatment or vaccine for DENV.

登革热是一种由登革热病毒引起蚊媒传染病，全世界约有3.9亿人受到感染，目前还没有针对登革热的有效疫苗。信封蛋白在病毒感染的膜融合过程中起到调节作用，来自于信封蛋白的肽抑制剂可以阻断膜融合这一过程，从而可能成为潜在的抗病毒药物。本项目拟设计以信封蛋白为靶点的抗登革热病毒的有效肽抑制剂。首先，在我们前期开发的蒙特卡洛算法中，使用统计势来评价信封蛋白的稳定性，加入信封蛋白的特异性对统计势进行优化会更有针对性。其次，在蒙特卡洛算法基础上加入遗传算法，可以对可能的抗登革热病毒肽抑制剂做进一步优化。再次，我们用支持向量机方法建立了有效肽抑制剂的物化性质和活性之间的关系模型，将模型和前期算法进行整合，通过模型进行肽抑制剂的选择更具有可靠性。最后，再通过生物学实验进行验证，主要包括病毒感染前后肽抑制剂对病毒的抑制实验和肽抑制剂对细胞的毒性检测实验。从而开发出抗登革热病毒的有效药物或疫苗的先导化合物。

登革热病毒严重威胁了人类的健康，目前仍无有效的治疗措施。膜融合是登革热病毒入侵细胞的第一步，信封蛋白的构象变化在膜融合过程中起到关键作用，因此研发针对信封蛋白的肽抑制剂为治疗登革热提供了潜在的重要靶点。本实验设计了优化针对该蛋白的肽抑制剂的遗传算法，分别使用三种不同的选择算子：轮盘法、锦标赛法和并行法，对初始父代进行300代的优化，通过比较发现并行法的结果较好，命名最优的结果为DN58ga，与此同时还获得了一些多肽衍生物，通过计算验证，这些多肽比父代表现出更好的与登革热病毒信封蛋白的结合能力。这些新的肽抑制剂有可能打断融合过程中的蛋白-蛋白相互作用，并有望作为开发登革热病毒信封蛋白新型抑制剂的起点。