病毒宿主和感染性预测及病毒系统分类的算法研究

针对此次新型冠状病毒的突出问题以及今后新发现病毒研究的重要需求，本项目致力于发展病毒宿主及其感染性预测、病毒系统分类的生物信息学新方法。基于不断丰富的病毒与宿主相关生物信息数据库，结合课题组自有的病毒相关的微生物组学数据，以数据为驱动，充分运用深度学习、自然语言处理的人工智能算法，突破当前基于序列相似性进行病毒宿主确定的局限，发展病毒宿主预测、病毒对宿主感染性预测的计算生物学新算法；基于病毒快速突变导致难以高效系统分类的挑战，发展对病毒进行大规模系统分类的的计算生物学新方法；进一步考虑面向直接测序获得的宏病毒组的微生物组数据，发展复杂微生物群落中目标病毒的宿主预测、病毒对宿主感染性预测，以及系统分类的计算生物学新方法；最终，搭建一个基于人工智能算法的新型冠状病毒及新发现病毒的宿主预测、感染性预测以及系统分类的计算生物学平台。

Addressing the important issues existing in the novel coronavirus and the demand research of newly discovered viruses in the future, this project is committed to the development of new methods of computational biology for the prediction of host and infectivity of newly discovered viruses and the systematic classification of them. Based on the information of viruses and their host contained in constantly enriched bioinformatics database, as well as the viral related microbiome data by our lab, we will take the full advantages of deep learning algorithms and natural language processing algorithms in artificial intelligence, to design new algorithms for predicting host and infectivity of virus, breaking through the current limitations of current sequence-similarity-based approaches of host identification, and to develop new methods of computational biology for large-scale systematic classification of viruses, of which the effectiveness was used to be limited by the rapid mutation of viruses. Furthermore, we will broaden the algorithms to the target virus in metavirome, which are obtained by direct sequencing of the virus community, predicting its host and infectivity and classifying it. Finally, we will build a computational biology platform for predicting the host and infectivity of new coronaviruses and other newly discovered viruses and classifying them systematically.