计算模型在化合物毒性预测方面的应用及展望

Application and prospects of computational models in predicting compound toxicity

毒性研究是一项重要且复杂的科学工作,涉及到评估化学物质及其他物质对人类健康的潜在危害,也对环境保护和可持续发展起着重要作用。传统的基于分子、细胞与动物模型的毒理学实验,除了成本高昂且耗时的问题,还存在对动物不友好的局限性。随着计算机技术的迅猛发展,出现了计算毒理学模型的开发研究。相比于传统的毒理学实验,计算毒理学模型具有高效率、低成本、方便快捷的特点,逐渐成为毒理学研究的热点方向之一。介绍了包括机器学习在内的计算模型预测化合物毒性的研究方法及其在环境保护、毒性风险评估和药物设计等领域的应用,并进一步讨论了计算模型在毒性预测中存在的不足之处及可能的发展方向。

Toxicity research is an important and complex scientific task that involves assessing the potential hazards of chemi⁃cals and other substances to human health,and plays an important role in environmental protection and sustainable development.Traditional toxicological experiments based on molecular,cellular,and animal models not only have high cost and time⁃consuming issues,but also have limitations in being unfriendly to animals.With the rapid development of computer technology,the research works of computational toxicology models have been emerged.Compared to traditional toxicology experiments,computational toxi⁃cology models have the characteristics of high efficiency,low cost,and convenience,gradually becoming one of the hot topics in toxicology research.This article introduces the research methods for predicting compound toxicity using computational models,in⁃cluding machine learning,and their applications in environmental protection,toxicity risk assessment,and drug design.It further discusses the shortcomings and potential development directions of computational models in toxicity prediction.