Data for Synthetic lethal connectivity and graph transformer improve synthetic lethality prediction

Here stores the data we used to train and evaluate our MLEC-iSL model for individual sample-specific synthetic lethality prediction. There are two types of features used in our model: omics features and biological networks. For omics features, there are two CCLE profiles (expression and essentiality) and four cell-specific features (expression, essentiality, mutation and copy number) considered in our model. As for network features, we incorporate physical protein-protein interaction (PPI) network, genetic interaction network and pathway network.本数据集存储了我们用于训练和评估针对个体样本特异性合成致死（synthetic lethality）预测任务的MLEC-iSL模型的相关数据。 本模型所采用的特征分为两类：组学（omics）特征与生物网络特征。其中组学特征包含两类CCLE（Cancer Cell Line Encyclopedia，癌症细胞系百科全书）谱图——表达谱与必需性谱，以及四类细胞特异性特征：表达、必需性、突变与拷贝数（copy number）特征。在网络特征方面，我们纳入了物理蛋白质-蛋白质相互作用（PPI）网络、遗传相互作用网络以及通路网络。