Search

Chemical reactions are the fundamental building blocks of drug design and organic chemistry research. In recent years, there has been a growing need for a large-scale deep-learning framework that can efficiently capture the basic rules of chemical reactions. In this paper, we have proposed a unified framework that addresses both the reaction representation learning and molecule generation tasks, which allows for a more holistic approach. Inspired by the organic chemistry mechanism, we develop a novel pretraining framework that enables us to incorporate inductive biases into the model. Our framework achieves state-of-the-art results on challenging downstream tasks. By possessing chemical knowledge, our generative framework overcome the limitations of current molecule generation models that rely on a small number of reaction templates. In the extensive experiments, our model generates synthesizable drug-like structures of high quality. Overall, our work presents a significant step toward a large-scale deep-learning framework for a variety of reaction-based applications.

Yiran Zhou,1,2,* Ruhong Li,2,* Haidong Yu,2 Ruotian Wang,2 Zhiqiang Shen1 1Department of Pharmacy, Kunming Medical University, 2Yan’an Hospital Affiliated to Kunming Medical University, Kunming, People’s Republic of China *These authors contributed equally to this work Abstract: Gastric cancer is one of the most common causes of death worldwide, although its incidence has steadily declined in recent years. There is strong evidence that aberrantly expressed microRNAs (miRNAs) are involved in gastric cancer tumorigenesis. Furthermore, CRMP4 is closely associated with the occurrence and development of gastric cancer, and our predictions suggest that miR-130a, which can promote gastric cancer tumorigenesis, is a potential CRMP4 regulator. In this study, we investigated the expression of CRMP4 and miR-130a in human gastric cancer cell lines by quantitative reverse transcription polymerase chain reaction (qRT-PCR) and Western blot (WB) examination and direct interactions between miR-130a and CRMP4 by dual-luciferase reporter assay. We also evaluated the biological roles of miR-130a and CRMP4 in gastric cancer cells by flow cytometry, MTT assay, soft agar colony formation assay, and Transwell tests and confirmed CRMP4 function in vivo, using a tumor xenograft model. Our results demonstrated that CRMP4 expression was significantly decreased at both the gene and protein levels, while miR-130a expression was notably increased, in five human gastric cancer cell lines compared with human gastric epithelial cells. Dual-luciferase reporter assays indicated that CRMP4 was the direct target of miR-130a. Moreover, an inverse regulatory relationship between miR-130a and CRMP4 was verified by qRT-PCR and WB, and overexpression of miR-130a in BGC823 cells enhanced apoptosis and cell proliferation, arrested the cell cycle in G0/G1, and facilitated cell colony formation, invasion, migration, and adhesion, while upregulation of CRMP4 had opposite effects. Fina