基于多维网络的癌症细胞系治疗药物预测方法 提出一种基于癌症细胞系相关的多层面数据预测药物反应的方法

The gene expression profile, somatic mutation, CNV, drug chemical structures and targets data were fused based on a multilayer network model and low-dimensional feature vector representation to predict drug response.  Firstly, a multilayer network including three cell line similarity networks and two drug similarity networks were constructed. Afterwards, a low-dimensional feature vector representation was used to fuse the biological information in the multilayer network. Then, a machine learning model, i.e. logistic regression(LR)  which was a successful model on classification problem of bioinformatics, was applied to predict new drug response according to known drug-cell lines associations.