The therapeutic efficacy of Danshen and Jiangxiang in the treatment of ischemic stroke (IS) is relatively significant. Studying the mechanism of action of Danshen and Jiangxiang in the treatment of IS can effectively identify candidate traditional Chinese medicines (TCM) with efficacy. However, it is challenging to analyze the effector substances and explain the mechanism of action of Danshen-Jiangxiang from a systematic perspective using traditional pharmacological approaches. In this study, a systematic study was conducted based on the drug-target-symptom-disease association network using complex network theory. On the basis of the association information about Danshen, Jiangxiang and IS, the protein-protein interaction (PPI) network and the “drug pair-pharmacodynamic ingredient-target-IS” network were constructed. The different topological features of the networks were analyzed to identify the core pharmacodynamic ingredients including formononetin in Jiangxiang, cryptotanshinone and tanshinone IIA in Danshen as well as core target proteins such as prostaglandin G/H synthase 2, retinoic acid receptor RXR-alpha, sodium channel protein type 5 subunit alpha, prostaglandin G/H synthase 1 and beta-2 adrenergic receptor. Further, a method for screening IS candidates based on TCM symptoms was proposed to identify key TCM symptoms and syndromes using the “drug pair-TCM symptom-syndrome-IS” network. The results showed that three TCMs, namely Puhuang, Sanleng and Zelan, might be potential therapeutic candidates for IS, which provided a theoretical reference for the development of drugs for the treatment of IS.
Protein structure determines function, and structural information is critical for predicting protein thermostability. This study proposes a novel method for protein thermostability prediction by integrating graph embedding features and network topological features. By constructing residue interaction networks (RINs) to characterize protein structures, we calculated network topological features and utilize deep neural networks (DNN) to mine inherent characteristics. Using DeepWalk and Node2vec algorithms, we obtained node embeddings and extracted graph embedding features through a TopN strategy combined with bidirectional long short-term memory (BiLSTM) networks. Additionally, we introduced the Doc2vec algorithm to replace the Word2vec module in graph embedding algorithms, generating graph embedding feature vector encodings. By employing an attention mechanism to fuse graph embedding features with network topological features, we constructed a high-precision prediction model, achieving 87.85% prediction accuracy on a bacterial protein dataset. Furthermore, we analyzed the differences in the contributions of network topological features in the model and the differences among various graph embedding methods, and found that the combination of DeepWalk features with Doc2vec and all topological features was crucial for the identification of thermostable proteins. This study provides a practical and effective new method for protein thermostability prediction, and at the same time offers theoretical guidance for exploring protein diversity, discovering new thermostable proteins, and the intelligent modification of mesophilic proteins.