文献类型：期刊文献

英文题名：Machine Learning-Based Multi-Modal and Multi-Granularity Feature Fusion Framework for Accurate Prediction of Molecular Properties

作者：Nan, Shihao[1];Li, Zhongmei[2];Jin, Saimeng[1];Du, Wenli[3];Shen, Weifeng[1]

机构：[1]Chongqing Univ, Sch Chem & Chem Engn, Chongqing 400044, Peoples R China;[2]East China Univ Sci & Technol, Key Lab Smart Mfg Energy Chem Proc, Minist Educ, Shanghai 200237, Peoples R China;[3]East China Univ Sci & Technol, State Key Lab Ind Control Technol, Minist Educ, Shanghai 200237, Peoples R China

年份：2025

卷号：64

期号：5

起止页码：3045

外文期刊名：INDUSTRIAL & ENGINEERING CHEMISTRY RESEARCH

收录：;EI(收录号：20250517774056);WOS:【SCI-EXPANDED(收录号:WOS:001403557300001)】；

基金：We acknowledge the financial support provided by the National Natural Science Foundation of China (nos. 22208036, 22278044); the National Natural Science Foundation for Excellent Young Scientists of China (no. 22122802); Venture & Innovation Support Program for Chongqing Overseas Returnees (CX2019117); the Chongqing Science Fund for Distinguished Young Scholars (no. CSTB2022NSCQ-JQX0021); the Chongqing Innovation Support Key Program for Returned Overseas Chinese Scholars (CX2023002).

语种：英文

摘要：The accurate prediction of molecular properties is a pivotal component in advancing chemical engineering technology. However, the efficacy of many machine learning-based quantitative structure-property relationship (QSPR) models is highly constrained by their reliance on specific types of molecular representations. To address this issue, a multi-modal and multi-granularity feature fusion framework has been designed to fully explore diverse information sources and improve the prediction accuracy of molecular properties. Initially, a self-supervised pretrained sequence feature encoder is developed, utilizing molecular fingerprints and atomic-level information to capture the intricate features of molecules. Meanwhile, the atom-level graph knowledge is integrated with the motif-level graph knowledge by developing a hierarchical graph feature encoder, and thereby enhancing the capacity to learn molecular topological information. Subsequently, several strategies including low-rank multimodal fusion are employed to synthesize the learned features. Comprehensive evaluations across four molecular data sets demonstrate that the proposed framework achieves superior accuracy and reliability. Through the analysis of the distribution of different features and comprehensive ablation studies, the ability of the proposed framework to capture multimodal features and extract additional potential information has been demonstrated. By systematically leveraging this information, the constructed framework enhances predictive performance and stability, thereby expanding the application prospects of machine learning-based QSPR in advancing intelligent chemical processes.