(EN) The invention discloses a method and device for screening COVID-19 drugs. On the basis of obtaining sequence similarity information between viruses and structure similarity information between pharmaceutical chemical structures, a heterogeneous network based on virus sequence similarity and drug similarity is constructed, and training is performed on the constructed data set by using a random walkalgorithm so as to obtain a COVID-19 drug screening model. Based on the drug screening model, drugs related to new coronal pneumonia can be effectively screened. According to the method, the researchand development cost of the COVID-19 drugs can be reduced, and the screening speed and accuracy of the COVID-19 drugs exceed those of representative methods applied to other bioinformatics fields.
(ZH) 本发明公开用于筛选新冠肺炎药物的方法及装置。本发明在获得病毒之间的序列相似信息和药物化学结构之间的结构相似信息基础上，通过构建基于病毒序列相似性和药物相似性的异构网络，并利用随机游走算法在构建的数据集上进行训练从而得到新冠肺炎药物筛选模型。基于该药物筛选模型能有效地筛选与新冠肺炎相关的药物。本发明的方法能够降低新冠肺炎药物研发成本，而且在新冠肺炎药物筛选的速度和准确率上均超过应用于其他生物信息学领域的代表性方法。