人工智能（AI）和機(jī)器學(xué)習(xí)不僅使藥物發(fā)現(xiàn)和開發(fā)實(shí)現(xiàn)了質(zhì)的飛躍，而且?guī)椭幬镩_發(fā)進(jìn)程進(jìn)入現(xiàn)代化。機(jī)器學(xué)習(xí)和深度學(xué)習(xí)算法已應(yīng)用于藥物發(fā)現(xiàn)各個(gè)階段，如先導(dǎo)化合物的篩選、多肽合成及小分子藥物的發(fā)現(xiàn)、最佳給藥劑量的確定、類藥化合物的設(shè)計(jì)和藥物不良反應(yīng)的預(yù)測、蛋白質(zhì)間相互作用的預(yù)測、虛擬篩選效率的提高、定量構(gòu)效關(guān)系（QSAR）建模和藥物重新定位、理化性質(zhì)和藥物靶標(biāo)親和力的預(yù)測、化合物的結(jié)合預(yù)測和體內(nèi)安全性分析、多靶點(diǎn)配體藥物分子的設(shè)計(jì)以及臨床試驗(yàn)的設(shè)計(jì)。簡要綜述了AI算法和傳統(tǒng)化學(xué)相結(jié)合以提高藥物發(fā)現(xiàn)的效率以及AI在藥物發(fā)現(xiàn)過程中的應(yīng)用研究進(jìn)展，以期為AI應(yīng)用于藥物發(fā)現(xiàn)提供一定參考。

Artificial intelligence (AI) and machine learning (ML) not only has made a qualitative leap in drug discovery and development, but also has helped the process of drug development enter modernization. Algorithms of ML and deep learning (DL) have been applied to various phases of drug discovery, such as screening of lead compounds, peptide synthesis and discovery of small molecular drugs, determination of the optimal dosage, design of drug-like compounds and prediction of adverse drug reaction (ADR), prediction of protein-protein interaction, improvement of the efficiency of virtual screening, quantitative structure-activity relationship (QSAR) modeling and drug repositioning, prediction of physicochemical properties and affinity of drug targets, prediction of drug binding and in vivo safety analysis, design of multiple target ligand drug molecules, as well as design of clinical trials. The research progress in the combination of AI algorithm and traditional chemistry to improve the efficiency of drug discovery and application of AI during the process of drug discovery was briefly reviewed, in order to provide some references for using AI for drug discovery in China.

中國食品藥品檢定研究院學(xué)科帶頭人課題（2021X2）