A complete graph-based approach with multi-task learning for predicting synergistic drug combinations.

MOTIVATION: Drug combination therapy shows significant advantages over monotherapy in cancer treatment. Since the combinational space is difficult to be traversed experimentally, identifying novel synergistic drug combinations based on computational methods has become a powerful tool for pre-screening. Among them, methods based on deep learning have far outperformed other methods. However, most deep learning-based methods are unstable and will give inconsistent predictions even by simply changing the input order of drugs. In addition, the insufficient experimental data of drug combination screening limits the generalization ability of existing models. These problems prevent the deep learning-based models from being in service. RESULTS: In this article, we propose CGMS to address the above problems. CGMS models a drug combination and a cell line as a heterogeneous complete graph, and generates the whole-graph embedding to characterize their interaction by leveraging the heterogeneous graph attention network. Based on the whole-graph embedding, CGMS can make a stable, order-independent prediction. To enhance the generalization ability of CGMS, we apply the multi-task learning technique to train the model on drug synergy prediction task and drug sensitivity prediction task simultaneously. We compare CGMS's generalization ability with six state-of-the-art methods on a public dataset, and CGMS significantly outperforms other methods in the leave-drug combination-out scenario, as well as in the leave-cell line-out and leave-drug-out scenarios. We further present the benefit of eliminating the order dependency and the discrimination power of whole-graph embeddings, interpret the rationality of the attention mechanism, and verify the contribution of multi-task learning. AVAILABILITY AND IMPLEMENTATION: The code of CGMS is available via https://github.com/TOJSSE-iData/CGMS.

Evaluation of cardiomyocyte hypoxia injury models for the pharmacological study in vitro.

CONTEXT: Traditional Herbal Medicine (THM) has many advantages that make it a promising choice for the treatment of ischemic heart disease (IHD). To study the mechanism of IHDs or pharmacological actions of THM, many hypoxia-induced cardiomyocyte injury models have been established. Radix Salvia miltorrhiza (Danshen) was used as a representative of THM. Danshen is a famous medicinal herb widely applied in Asia to relieve ischemic cardiovascular diseases. OBJECTIVE: To investigate the effects of various hypoxic conditions and discuss a suitable hypoxia model, cell viability, apoptosis, release of myocardial injury markers, and mRNA levels of target genes were tested for the first time. MATERIALS AND METHODS: Radix Salvia miltorrhiza (Danshen) was purchased from a GMP-compliant producer and both its preparation method and quality control were standardized. Cellular status, such as cell viability, apoptosis, releases of myocardial injury markers, and the mRNA level of target gene were tested by 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) method, biochemical analyzer, flow cytometry, Hoechst 33258 staining, and real-time PCR, respectively. RESULTS: Based on our data, we found a treppe response of cardiomyocyte in the hypoxic condition and suggested that 8 h in 2% O2 might be a suitable condition for in vitro pharmacological study of cardiomyocytes. DISCUSSION AND CONCLUSIONS: Our findings outlined more extended and in-depth capability of cardiomyocyte suffering from hypoxia, and might be of particular interest due to the high prevalence of THM pharmacological study.