Qingyao Ding,1 Rufan Yao,1 Yue Bai,1 Limu Da,1 Yujiang Wang,2 Rongwu Xiang,1,3,4 Xiwei Jiang,1 Fei Zhai1 

1Faculty of Medical Devices, Shenyang Pharmaceutical University, Shenyang, Liaoning Province, People’s Republic of China; 2Department of Internal Medicine, Zhengding County People’s Hospital, Shijiazhuang, Hebei Province, People’s Republic of China; 3Liaoning Medical Big Data and Artificial Intelligence Engineering Technology Research Center, Shenyang, Liaoning Province, People’s Republic of China; 4Institute of Regulatory Science for Medical Products, Shenyang Pharmaceutical University, Shenyang, Liaoning Province, People’s Republic of China

Correspondence: Xiwei Jiang, Faculty of Medical Devices, Shenyang Pharmaceutical University, Shenyang, 110016, People’s Republic of China, Email jiangxiwei@syphu.edu.cn Fei Zhai, Faculty of Medical Devices, Shenyang Pharmaceutical University, Shenyang, 110016, People’s Republic of China, Email 106030309@syphu.edu.cn

Abstract: In recent years, the widespread use of artificial intelligence (AI) and big data technologies in drug research has significantly accelerated the drug development process. However, their black-box nature makes it challenging to evaluate their effectiveness and safety. The interpretability of models has become a key issue in the application of AI in the drug development. In this paper, a bibliometric approach has been adopted to systematically analyze the application of Explainable Artificial Intelligence (XAI) techniques in drug research, with an in-depth discussion of the developmental trends, geographical distribution, journal preferences, major contributors, and research hotspots. In addition, the research results of XAI are summarized in the three directions of chemical, biological, and traditional Chinese medicine, and the future research directions and development trends are envisioned in order to promote the in-depth application of XAI technology in drug discovery and continuous innovation.

Keywords: explainable artificial intelligence, XAI, drug research, bibliometric analysis, interpretability, shapley additive explanations, SHAP