流感病毒一直以来是人类健康的重要威胁，其传染性和流行性往往会造成大面积人群的感染，不仅影响社会稳定与经济发展，也对国家生物安全造成严重威胁。目前，接种疫苗是流感病毒防治最为有效的策略之一。然而，由于流感病毒易产生快速突变，通过实验手段筛选得到的流感病毒疫苗存在免疫反应弱，覆盖范围窄，易失效等缺陷。本项目拟针对流感病毒重要免疫原血凝素蛋白，构建一套广谱性疫苗计算设计流程，结合实验验证筛选得到流感病毒广谱疫苗。本项目通过构建血凝素蛋白的序列数据及蛋白结构数据集，利用特征筛选，机器学习，数理模型构建等手段构建血凝素蛋白免疫原性区域预测模型和抗原性相似性预测模型，并结合实验验证对模型进行反馈修订。基于本模型的高通量免疫原性评估与大规模抗原性保护谱构建，能够缩短疫苗筛选进程，提高疫苗有效性，为流感病毒的疫苗免疫原设计提供理论基础和方向指导。

Influenza virus is a continuous threat to human health. Its infectivity and epidemiology often cause large number of infections in the community, which not only affects social stability and economic development, but also poses a serious threat to national bio-safety. Currently, vaccination is one of the most effective strategies for influenza virus control and treatment. However, due to the rapid mutation of influenza virus, the vaccine screened by experiment has some shortcomings, such as weak immune response, narrow protective ability and recurrent failure. Here, we aim to construct a computational pipeline for broad-spectrum vaccine design based on the hemagglutinin proteins of influenza virus, then test the utility of this pipeline with experimental data. In this project, the sequence data and protein structure data of hemagglutinin protein is collected as our dataset. Then, a computational model is established for immunogenicity and antigenicity prediction, based on the feature selecting, machine learning approaches and mathematical models. High-throughput immunogenicity and large-scale antigenicity prediction based on this project can shorten the process of vaccine screening, improve the vaccine effectiveness, and further, provide theoretical basis and direction guidance of vaccine immunogenic design for influenza virus.