AdapGNN: enhancing the explainability of GNN models in molecular properties prediction

Abstract

Graph neural networks (GNNs) have played an increasingly important role in molecular property prediction. However, GNN models are prone to face the oversmoothing problem. By reviewing existing works on addressing oversmoothing, we noticed that those methods are designed for graph data and often break the original topological structure. However, it is not acceptable in molecule data which the real physicochemical meanings are lost. Motivated by this, to address the problem of oversmoothing and fill the gap in molecule property prediction, we proposed AdapGNN, a novel model-agnostic framework that designed for molecule property prediction specially. This is achieved through the integration of original node feature into the message-passing step of GNN models. Besides, to emphasize the crucial part of a molecule during predicting and further enhance the explainability of our model, we proposed a weight projection module to generate node-specific weight when merging node features. Furthermore, to validate the efficiency of our method and addressing the problem that existing benchmark data set lacks of the ground-truth of atom importance. We proposed MolExplain, a new benchmark data set for quantitative explainability evaluation in molecule property prediction. Experimental results show that the AdapGNN significantly improves the explainability of GNN models while maintaining high predictive accuracy.

Graphical abstract

Keywords

GNN / oversmoothing / explainability / molecular property prediction / benchmark

References

[1] Liu Z , Du J , Fang J , Yin Y , Xu G , Xie L . DeepScreening: a deep learning-based screening web server for accelerating drug discovery. Database, 2019, 2019: baz104

[2] Cai C , Wang S , Xu Y , Zhang W , Tang K , Ouyang Q , Lai L , Pei J . Transfer learning for drug discovery. Journal of Medicinal Chemistry, 2020, 63(16): 8683–8694

[3] Li P , Wang J , Qiao Y , Chen H , Yu Y , Yao X , Gao P , Xie G , Song S . An effective self-supervised framework for learning expressive molecular global representations to drug discovery. Briefings in Bioinformatics, 2021, 22(6): bbab109

[4] Lu X , Xie L , Xu L , Mao R , Xu X , Chang S . Multimodal fused deep learning for drug property prediction: integrating chemical language and molecular graph. Computational and Structural Biotechnology Journal, 2024, 23: 1666–1679

[5] Mercado R , Rastemo T , Lindelöf E , Klambauer G , Engkvist O , Chen H , Bjerrum E J . Graph networks for molecular design. Machine Learning: Science and Technology, 2021, 2(2): 025023

[6] Shi X , Cheng D , Zhao R , Zhang G , Wu S , Zhen S , Zhao Z J , Gong J . Accessing complex reconstructed material structures with hybrid global optimization accelerated via on-the-fly machine learning. Chemical Science, 2023, 14(33): 8777–8784

[7] Zhang H , Liao L , Saravanan K , Yin P , Wei Y . DeepBindRG: a deep learning based method for estimating effective protein-ligand affinity. PeerJ, 2019, 7: e7362

[8] Lin H , Zhu J , Wang S , Li Y , Pei J , Lai L . DeepRLI: a multi-objective framework for universal protein-ligand interaction prediction. Digital Discovery, 2025, 4(8): 2083–2103

[9] Hao ZLu CHuang ZWang HHu ZLiu QChen ELee C. ASGN: an active semi-supervised graph neural network for molecular property prediction. In: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. New York: Association for Computing Machinery, 2020, 731–752

[10] Zhang Z , Guan J , Zhou S . FraGAT: a fragment-oriented multi-scale graph attention model for molecular property prediction. Bioinformatics, 2021, 37(18): 2981–2987

[11] Xu S , Liu X , Cai P , Li J , Wang X , Liu B . Machine-learning-assisted accurate prediction of molecular optical properties upon aggregation. Advanced Science, 2022, 9(2): 2101074

[12] Lu S , Gao Z , He D , Zhang L , Ke G . Data-driven quantum chemical property prediction leveraging 3D conformations with Uni-Mol+. Nature Communications, 2024, 15(1): 7104

[13] Jow H , Sprung J , Ritchie L . et al. MELCOR Accident Consequence Code System (MACCS). , 1990,

[14] Schwartz J , Awale M , Reymond J L . SMIfp (SMILES fingerprint) chemical space for virtual screening and visualization of large databases of organic molecules. Journal of Chemical Information and Modeling, 2013, 53(8): 1979–1989

[15] Morgan H L . The generation of a unique machine description for chemical structures-a technique developed at chemical abstracts service. Journal of Chemical Documentation, 1965, 5(2): 107–113

[16] Rogers D , Hahn M . Extended-connectivity fingerprints. Journal of Chemical Information and Modeling, 2010, 50(5): 742–754

[17] Carhart R E , Smith D H , Venkataraghavan R . Atom pairs as molecular features in structure-activity studies: definition and applications. Journal of Chemical Information and Computer Sciences, 1985, 25(2): 64–73

[18] Nilakantan R , Bauman N , Dixon J S , Venkataraghavan R . Topological torsion: a new molecular descriptor for SAR applications, comparison with other descriptors. Journal of Chemical Information and Computer Sciences, 1987, 27(2): 82–85

[19] Kearnes S , McCloskey K , Berndl M , Pande V , Riley P . Molecular graph convolutions: moving beyond fingerprints. Journal of Computer-Aided Molecular Design, 2016, 30(8): 595–608

[20] Gilmer JSchoenholz S SRiley P FVinyals ODahl G E. Neural message passing for quantum chemistry. In: Proceedings of the 34th International Conference on Machine Learning. Sydney: Proceedings of Machine Learning Research, 2017, 1263–1272

[21] Jiang BZhang ZLin DTang JLuo B. Semi-supervised learning with graph learning-convolutional networks. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Los Alamitos: IEEE Computer Society, 2019, 11313–11320

[22] Xiong Z , Wang D , Liu X , Zhong F , Wan X , Li X , Li Z , Luo X , Chen K , Jiang H . et al. Pushing the boundaries of molecular representation for drug discovery with the graph attention mechanism. Journal of Medicinal Chemistry, 2020, 63(16): 8749–8760

[23] Zhang XXu YHe WGuo WCui L. A comprehensive review of the oversmoothing in graph neural networks. In: CCF Conference on Computer Supported Cooperative Work and Social Computing. Harbin: Springer, 2023, 451–465

[24] Rong Y , Huang W , Xu T , Huang J . Dropedge: towards deep graph convolutional networks on node classification. , 2019, arXiv:1907.10903:

[25] Papp P A , Martinkus K , Faber L , Wattenhofer R . DropGNN: Random dropouts increase the expressiveness of graph neural networks.. , 2021, arXiv:2111.06283:

[26] Hssayni E , Boufssasse A , Joudar N E , Ettaouil M . Novel dropout approach for mitigating over-smoothing in graph neural networks. Applied Intelligence, 2025, 55(5): 354

[27] Xu KLi CTian YSonobe TKawarabayashi K IJegelka S. Representation learning on graphs with jumping knowledge networks. In: International Conference on Machine Learning. Stockholm: Proceedings of Machine Learning Research, 2018, 5453–5462

[28] Hamilton W , Ying Z , Leskovec J . Inductive representation learning on large graphs. In: Advances in Neural Information Processing Systems. California: Curran Associates, 2017,

[29] Kitsupapaisan J , Maneeroj S , Takasu A . Integrating a sequential model into GNN-based social recommendation for relieving over-smoothing problem. IEEE Access: Practical Innovations, Open Solutions, 2025, 13: 100219–100233

[30] Zheng CZong BCheng WSong DNi JYu WChen HWang W. Robust graph representation learning via neural sparsification. In: International Conference on Machine Learning. San Diego: Proceedings of Machine Learning Research, 2020, 11458–11468

[31] Chen D , Lin Y , Li W , Li P , Zhou J , Sun X . Measuring and relieving the over-smoothing problem for graph neural networks from the topological view. Proceedings of the AAAI Conference on Artificial Intelligence, 2020, 34(4): 3438–3445

[32] Hossain TSaifuddin K MIslam M I KTanvir FAkbas E. Tackling oversmoothing in GNN via graph sparsification. In: Joint European Conference on Machine Learning and Knowledge Discovery in Databases. Cham: Springer Nature Switzerland, 2024, 161–179

[33] Wu Z , Ramsundar B , Feinberg E N , Gomes J , Geniesse C , Pappu A S , Leswing K , Pande V . MoleculeNet: a benchmark for molecular machine learning. Chemical Science, 2018, 9(2): 513–530

[34] Selvaraju R RCogswell MDas AVedantam RParikh DBatra D. Grad-cam: visual explanations from deep networks via gradient-based localization. In: Proceedings of the IEEE International Conference on Computer Vision. Cambridge: Institute of Electrical and Electronics Engineers (IEEE), 2017, 618–626

[35] Luo D , Cheng W , Xu D , Yu W , Zong B , Chen H , Zhang X . Parameterized explainer for graph neural network. Advances in Neural Information Processing Systems, 2020, 33: 19620–19631

[36] Huang Q , Yamada M , Tian Y , Singh D , Chang Y . GraphLIME: local interpretable model explanations for graph neural networks. IEEE Transactions on Knowledge and Data Engineering, 2023, 35(7): 6968–6972

[37] Yuan HTang JHu XJi S. Xgnn: towards model-level explanations of graph neural networks. In: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. San Diego: Association for Computing Machinery, 2020, 430–438

[38] Ying R , Bourgeois D , You J , Zitnik M , Leskovec J . GNNExplainer: generating explanations for graph neural networks. Advances in Neural Information Processing Systems, 2019, 32: 9240–9251

[39] Mendez D , Gaulton A , Bento A P , Chambers J , De Veij M , Félix E , Magariños M P , Mosquera J F , Mutowo P , Nowotka M . et al. ChEMBL: towards direct deposition of bioassay data. Nucleic Acids Research, 2019, 47(D1): D930–D940

RIGHTS & PERMISSIONS

Higher Education Press