Skip to content
2000
Volume 20, Issue 9
  • ISSN: 1574-8936
  • E-ISSN: 2212-392X

Abstract

Introduction/ Objective

Graph neural networks’ (GNNs) explainability, especially the explanation of edges and interactions among vertices in GNNs, is demanding mainly owing to dynamics and groupings between vertices. The existing graph explainability methods ignore the analysis of the following tasks weights over subgraphs but instead analyze solely sample-level explainability. Such sample-level explainability decreases their generalizability since it directly searches the explaining behaviour in the input dataset. In this study, we come up with a novel Orbit-based GNN explainer (OExplainer), which integrates both sample-level and method-level approaches over a predetermined set of subgraphs. As part of such analysis of subgraphs, our goal is to interpret graphs more comprehensively and intelligibly while providing each vertex’s explainability score for a particular graph instance.

Methods

Our OExplainer decomposes the following graph neural network weights into explaining subgraph bases while identifying and characterizing particular predictions. By such characterization, we can carefully and accurately interpret the predetermined graph orbit’s role in vertex representation determination. In this characterization, we can also clarify the method’s behaviour generally for the whole input dataset. Moreover, we come up with novel vertex-specific scores in our subgraph-based approach over nonisomorphic graphlets. Such vertex-specific score encourages sample-level vertex improvement, and such improvement is related to the graph neural network’s vertex classification task.

Results

Our experiments over simulated datasets confirm the importance and criticality of method weights in vertex classification explanation. In this case, method weight decomposition also has criticality. Our detailed experiments over multiple real protein-protein interaction datasets and metabolic interaction networks also exhibit enhanced performance in vertex classification.

Conclusion

In both simulated and biological protein-protein interaction datasets, our approach outperforms the competing explanation approaches.

© 2025 Bentham Science Publishers
Loading

Article metrics loading...

/content/journals/cbio/10.2174/0115748936355418250114104026
2025-02-06
2025-10-27

  • From This Site

    /content/journals/cbio/10.2174/0115748936355418250114104026
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. HoganA. BlomqvistE. CochezM. D’amatoC. MeloG.D. GutierrezC. KirraneS. GayoJ.E.L. NavigliR. NeumaierS. NgomoA-C.N. PolleresA. RashidS.M. RulaA. SchmelzeisenL. SequedaJ. StaabS. ZimmermannA. Knowledge graphs.ACM Comput. Surv.202254413710.1145/3447772
     [Google Scholar]
  2. SeferE. BioCode: A data-driven procedure to learn the growth of biological networks.IEEE/ACM Trans. Comput. Biol. Bioinf.20221963103311310.1109/TCBB.2022.316509235380966
     [Google Scholar]
  3. DuvenaudD. MaclaurinD. IparraguirreA.J. BombarelliG.R. HirzelT. GuzikA.A. Convolutional networks on graphs for learning molecular fingerprints.proceedings of the 28th International Conference on Neural Information Processing SystemsCambridge, MA, USA MIT Press, 2015, pp. 2224-2232.
     [Google Scholar]
  4. SeferE. ProbC: Joint modeling of epigenome and transcriptome effects in 3D genome.BMC Genomics202223128710.1186/s12864‑022‑08498‑535397520
     [Google Scholar]
  5. SeferE. Hi–C interaction graph analysis reveals the impact of histone modifications in chromatin shape.Appl. Netw. Sci.2021615410.1007/s41109‑021‑00396‑1
     [Google Scholar]
  6. XingZ. ZhaoS. GuoW. MengF. GuoX. WangS. HeH. Coal resources under carbon peak: Segmentation of massive laser point clouds for coal mining in underground dusty environments using integrated graph deep learning model.Energy202328512877110.1016/j.energy.2023.128771
     [Google Scholar]
  7. FanW. MaY. LiQ. HeY. ZhaoE. TangJ. Graph neural networks for social recommendation.The World Wide Web Conference. WWW ’19New York, NY, USA, 2019, pp. 417-426.10.1145/3308558.3313488
     [Google Scholar]
  8. PatroR. SeferE. MalinJ. MarçaisG. NavlakhaS. KingsfordC. Parsimonious reconstruction of network evolution.Algorithms in Bioinformatics. PrzytyckaT.M. SagotM.F. Berlin, HeidelbergSpringer Berlin Heidelberg201123724910.1007/978‑3‑642‑23038‑7_21
     [Google Scholar]
  9. SeferE. KingsfordC. Metric Labeling and Semi-metric Embedding for Protein Annotation Prediction.Research Complement Molecular Biology BafnaV. SahinalpS.C. Berlin, HeidelbergSpringer Berlin Heidelberg201139240710.1007/978‑3‑642‑20036‑6_37
     [Google Scholar]
  10. HamiltonW.L. YingR. LeskovecJ. Inductive representation learning on large graphs.Proceedings of the 31st International Conference on Neural Information Processing Systems. NIPS’17Red Hook, NY, USA, 2017, pp. 1025-1035.
     [Google Scholar]
  11. GeziciA.H.B. SeferE. Deep transformer-based asset price and direction prediction.IEEE Access202412241642417810.1109/ACCESS.2024.3358452
     [Google Scholar]
  12. PatroR. DuggalG. SeferE. WangH. FilippovaD. KingsfordC. The missing models: A data-driven approach for learning how networks grow.Grow. Model.20121910.1145/2339530.2339541
     [Google Scholar]
  13. KipfT.N. WellingM. Semi supervised classification with graph convolutional networks.5th International Conference on Learning Representations, ICLR 2017Toulon, France, 2017, pp. 1-6.10.48550/arXiv.1609.02907
     [Google Scholar]
  14. VelickovicP. CucurullG. CasanovaA. RomeroA. LiòP. BengioY. Graph attention networks.Mach. Learn.201731090310.48550/arXiv.1710.10903
     [Google Scholar]
  15. XuK. HuW. LeskovecJ. JegelkaS. How powerful are graph neural networks?arXiv:1810.00826.201810.48550/arXiv.1810.00826
     [Google Scholar]
  16. WuY. ZhaoS. XingZ. WeiZ. LiY. LiY. Detection of foreign objects intrusion into transmission lines using diverse generation model.IEEE Trans. Power Deliv.20233853551356010.1109/TPWRD.2023.3279891
     [Google Scholar]
  17. YingZ. BourgeoisD. YouJ. ZitnikM. LeskovecJ. GNNExplainer: Generating Explanations for Graph Neural Networks.33rd Conference on Neural Information Processing Systems (NeurIPS 2019)Vancouver, Canada, 2019, pp. 1-12.
     [Google Scholar]
  18. LuoD. ChengW. XuD. YuW. ZongB. ChenH. Parameterized explainer for graph neural network.Proceedings of the 34th International Conference on Neural Information Processing Systems. NIPS’20Red Hook, NY, USA, 2020, p. 04573.10.48550/arXiv.2011.04573
     [Google Scholar]
  19. YuZ. GaoH. MotifExplainer: A motif-based graph neural network explainer.ArXiv abs/2202.005192022
     [Google Scholar]
  20. YuanH. YuH. WangJ. LiK. JiS. On explainability of graph neural networks via subgraph explorations.Proceedings of Machine Learning Research MeilaM. ZhangT. Proceedings of the 38th International Conference on Machine Learning.202113912224112252
     [Google Scholar]
  21. ShinY.M. KimS.W. ShinW.Y. PAGE: Prototype-based model-level explanations for graph neural networks.arXiv:2210.1715920221610.48550/arXiv.2210.17159
     [Google Scholar]
  22. YuanH. TangJ. HuX. JiS. XGNN: Towards Model-Level Explanations of Graph Neural Networks.Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. KDD ’20New York, NY, USA, 2020 pp. 430-438.10.1145/3394486.3403085
     [Google Scholar]
  23. AhmedN.K. NevilleJ. RossiR.A. DuffieldN. Efficient Graphlet Counting for Large Networks.2015 IEEE International Conference on Data MiningAtlantic City, NJ, USA, 2015, pp. 1-10.
     [Google Scholar]
  24. ChenX. LuiJ.C.S. Mining Graphlet Counts in Online Social Networks.2016 IEEE 16th International Conference on Data Mining (ICDM)Barcelona, Spain, 2016, pp. 71-80.10.1109/ICDM.2016.0018
     [Google Scholar]
  25. PržuljN. Biological network comparison using graphlet degree distribution.Bioinformatics200701232e177e18310.1093/bioinformatics/btl301
     [Google Scholar]
  26. PržuljN. CorneilDG. JurisicaI. Modeling interactome: Scale-free or geometric?Bioinformatics20040720183508351510.1093/bioinformatics/bth436
     [Google Scholar]
  27. KarpP.D. BillingtonR. CaspiR. FulcherC.A. LatendresseM. KothariA. KeselerI.M. KrummenackerM. MidfordP.E. OngQ. OngW.K. PaleyS.M. SubhravetiP. The BioCyc collection of microbial genomes and metabolic pathways.Brief. Bioinform.20192041085109310.1093/bib/bbx08529447345
     [Google Scholar]
  28. MarcusD. ShavittY. RAGE – A rapid graphlet enumerator for large networks.Comp. Net.201256281081910.1016/j.comnet.2011.08.019
     [Google Scholar]
  29. MawhirterD. WuB. MehtaD. AiC. An Approximate Graphlet Counting Algorithm for Large Graph Analysis.Proceedings of the 21st ACM International Conference on Information and Knowledge Management. CIKM ’12.New York, NY, USA, 2018, pp. 1467-1471.
     [Google Scholar]
  30. RahmanM. BhuiyanM. HasanM.A. Context-dependent reorganization of spatial and movement representations by simultaneously recorded hippocampal and striatal neurons during performance of allocentric and egocentric tasksBehav Neurosci.201211847519610.1145/2396761.239845415301602
     [Google Scholar]
  31. HocˇevarT. DemšarJ. A combinatorial approach to graphlet countingBioinformatics20141230455956510.1093/bioinformatics/btt717
     [Google Scholar]
  32. FortinS. The Graph Isomorphism ProblemSys. Net.19966311128134
     [Google Scholar]
  33. JansonS. LuczakT. RucinskiA. Random graphs.John Wiley & SonsHoboken, New Jersey201116
     [Google Scholar]
  34. DurrettR. Random graph dynamics.Cambridge university pressDuke University, North Carolina20101910.1017/CBO9780511546594
     [Google Scholar]
  35. NewmanM.E.J. The structure and function of complex networks.SIAM Rev.200345216725610.1137/S003614450342480
     [Google Scholar]
  36. ErdosP. RényiA. On the evolution of random graphsPubl. Math. Inst. Hung. Acad. Sci.1960511760
     [Google Scholar]
  37. BarabásiAL. AlbertR. Emergence of scaling in random networksScience1999286543950951210.1126/science.286.5439.50910521342
     [Google Scholar]
  38. AlbertR. BarabásiA.L. Statistical mechanics of complex networks.Rev. Mod. Phys.2002741479710.1103/RevModPhys.74.47
     [Google Scholar]
  39. DorogovtsevS.N. MendesJ.F.F. Evolution of networks.Adv. Phys.20025141079118710.1080/00018730110112519
     [Google Scholar]
  40. HolmeP. KimB.J. Growing scale-free networks with tunable clustering.Phys. Rev. E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Topics200265202610710.1103/PhysRevE.65.02610711863587
     [Google Scholar]
  41. ZitnikM. LeskovecJ. Predicting multicellular function through multi-layer tissue networksBioinformatics2017073314i190i19810.1093/bioinformatics/btx252
     [Google Scholar]
  42. GreeneC.S. KrishnanA. WongA.K. RicciottiE. ZelayaR.A. HimmelsteinD.S. ZhangR. HartmannB.M. ZaslavskyE. SealfonS.C. ChasmanD.I. FitzGeraldG.A. DolinskiK. GrosserT. TroyanskayaO.G. Understanding multicellular function and disease with human tissue-specific networks.Nat. Genet.201547656957610.1038/ng.325925915600
     [Google Scholar]
  43. AshburnerM. BallC.A. BlakeJ.A. BotsteinD. ButlerH. CherryJ.M. DavisA.P. DolinskiK. DwightS.S. EppigJ.T. HarrisM.A. HillD.P. TarverI.L. KasarskisA. LewisS. MateseJ.C. RichardsonJ.E. RingwaldM. RubinG.M. SherlockG. Gene Ontology: Tool for the unification of biology.Nat. Genet.2000251252910.1038/7555610802651
     [Google Scholar]
  44. YuanH. YuH GuiS JiS. Explainability in graph neural networks: A taxonomic survey.IEEE Trans Pattern Anal Mach Intell202245557829910.1109/TPAMI.2022.320423636063508
     [Google Scholar]
  45. KanehisaM. The KEGG database.Novartis Found Symp20022479110110.1002/0470857897.ch8
     [Google Scholar]
  46. MorrisC. RitzertM. FeyM. HamiltonW.L. LenssenJ.E. Weisfeiler and Leman Go Neural: Higher-Order Graph Neural Networks.Proceedings of the thirty-third aaai conference on artificial intelligence and thirty-first innovative applications of artificial intelligence conference and ninth aaai symposium on educational advances in artificial intelligence. AAAI’19/IAAI’19/EAAI’19.AAAI Press, 2019, pp. 1-6.10.1609/aaai.v33i01.33014602
     [Google Scholar]
  47. XieY. KatariyaS. TangX. HuangE. RaoN. SubbianK. Task-agnostic graph explanations.arXiv:2202.08335202216
     [Google Scholar]
/content/journals/cbio/10.2174/0115748936355418250114104026
Loading
A Graphlet-based Explanation Generator for Graph Neural Networks Over Biological Datasets
Curr Bioinform 20, 840 (2025); https://doi.org/10.2174/0115748936355418250114104026
/content/journals/cbio/10.2174/0115748936355418250114104026
/content/journals/cbio/10.2174/0115748936355418250114104026
Loading

Data & Media loading...


  • Article Type:     Research Article
Keyword(s): Deep learning; GNNs; interpretability; metabolic network; OExplainer; protein-protein interaction network

Most Cited Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error
Please enter a valid_number test