RESUMEN
To explore the therapeutic effects along with the molecular mechanisms of epigallocatechin gallate (EGCG) in non-alcoholic fatty liver disease (NAFLD) treatment using network pharmacology as well as animal experiments. Firstly, the Traditional Chinese Medicine (TCM) Systems Pharmacology Database was searched to identify the potential targets of EGCG. The DisGeNET Database was used to screen the potential targets of NAFLD. The GeneCards Database was searched to identify related genes involved in pyroptosis. Subsequently, the intersecting genes of EGCG targeting pyroptosis to regulate NAFLD were obtained using a Venn diagram. Simultaneously, the aforementioned intersecting genes were used to construct a drug-disease target protein-protein interaction (PPI) network. The DAVID database was adopted for Gene Ontology (GO) as well as Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. The main pathway-target network was determined. Next, the potential mechanism of EGCG targeting pyroptosis to regulate NAFLD was investigated and validated through in vivo experiments. 626 potential targets of EGCG, 447 target genes of NAFLD, and 568 potential targets of pyroptosis were identified. The number of common targets between EGCG, NAFLD, and pyroptosis was 266. GO biological process items and 92 KEGG pathways were determined based on the analysis results. Animal experiments demonstrated that EGCG could ameliorate body weight, glucolipid metabolism, steatosis, and liver injury, enhance insulin sensitivity, and improve glucose tolerance in NAFLD mice through the classical pathway of pyroptosis. EGCG could effectively treat NAFLD through multiple targets and pathways. It was concluded that EGCG ameliorates hepatocyte steatosis, pyroptosis, dyslipidemia, and inflammation in NAFLD mice fed a high-fat diet (HFD), and the protective mechanism could be associated with the NLRP3-Caspase-1-GSDMD classical pyroptosis pathway.
RESUMEN
Molecular property prediction is an essential but challenging task in drug discovery. The recurrent neural network (RNN) and Transformer are the mainstream methods for sequence modeling, and both have been successfully applied independently for molecular property prediction. As the local information and global information of molecules are very important for molecular properties, we aim to integrate the bi-directional gated recurrent unit (BiGRU) into the original Transformer encoder, together with self-attention to better capture local and global molecular information simultaneously. To this end, we propose the TranGRU approach, which encodes the local and global information of molecules by using the BiGRU and self-attention, respectively. Then, we use a gate mechanism to reasonably fuse the two molecular representations. In this way, we enhance the ability of the proposed model to encode both local and global molecular information. Compared to the baselines and state-of-the-art methods when treating each task as a single-task classification on Tox21, the proposed approach outperforms the baselines on 9 out of 12 tasks and state-of-the-art methods on 5 out of 12 tasks. TranGRU also obtains the best ROC-AUC scores on BBBP, FDA, LogP, and Tox21 (multitask classification) and has a comparable performance on ToxCast, BACE, and ecoli. On the whole, TranGRU achieves better performance for molecular property prediction. The source code is available in GitHub: https://github.com/Jiangjing0122/TranGRU.