The precise prediction of molecular properties is essential for advancements in drug development, particularly in virtual screening and compound optimization. The recent introduction of numerous deep learning-based methods has shown remarkable potential in enhancing molecular property prediction (MPP), especially improving accuracy and insights into molecular structures. Yet, two critical questions arise: does the integration of domain knowledge augment the accuracy of molecular property prediction and does employing multi-modal data fusion yield more precise results than unique data source methods? To explore these matters, we comprehensively review and quantitatively analyze recent deep learning methods based on various benchmarks. We discover that integrating molecular information will improve both MPP regression and classification tasks by upto 3.98% and 1.72%, respectively. We also discover that the utilizing 3-dimensional information with 1-dimensional and 2-dimensional information simultaneously can substantially enhance MPP upto 4.2%. The two consolidated insights offer crucial guidance for future advancements in drug discovery.

翻译：分子性质的精确预测对于药物研发的进步至关重要，尤其是在虚拟筛选和化合物优化方面。近年来，大量基于深度学习的方法在提升分子性质预测（MPP）方面展现出显著潜力，尤其是在提高准确性及对分子结构的洞察上。然而，两个关键问题随之浮现：领域知识的整合是否能增强分子性质预测的准确性？采用多模态数据融合是否比单一数据源方法能产生更精确的结果？为探究这些问题，我们基于多种基准对近期深度学习方法进行了全面综述和定量分析。我们发现，整合分子信息可分别将MPP回归和分类任务提升多达3.98%和1.72%。我们还发现，同时利用三维信息与一维、二维信息可显著提升MPP性能高达4.2%。这两项综合见解为未来药物发现的发展提供了重要指导。