基于GAT双聚合运算与归纳式矩阵补全的关联预测

doi:10.19678/j.issn.1000-3428.0063069

摘要/Abstract

摘要： 可计算模型能够有效替代生物实验进行长链非编码RNA（lncRNA）-疾病的关联预测，但由于存在已知数据稀疏性问题，导致现有模型的预测精度不高。针对这一局限性，提出基于图注意力网络与归纳式矩阵补全技术的双融合机制lncRNA-疾病关联预测模型（DFMP-LDA）。引入n头注意力机制，设计带有双重聚合器的图注意力网络，增强lncRNA节点与疾病节点的特征，避免数据稀疏性导致模型预测精度不高的问题。在此基础上，针对传统图注意力网络不能直接应用于潜在lncRNA-疾病对关联预测的问题，引入归纳式矩阵补全技术，应用增强后的节点特征重建lncRNA-疾病关联网络，进一步提高模型的预测精度。5折交叉验证结果表明，DFMP-LDA预测lncRNA-疾病关联的AUC值为0.932 2，AUPR值为0.770 5，在时间成本上分别较DMF-LDA、SDLDA、TPGLDA模型节省33.89%、32.17%、16.12%，预测性能较优。

关键词: 图注意力网络, 归纳式矩阵补全, 关联预测, 双重聚合器, 特征增强

Abstract: Computational models have been applied in long non-coding RNA(lncRNA)-disease association prediction to effectively replace traditional biological experiments.Due to the sparsity lack of input data, however, the prediction accuracy of existing models remains low.To address this limitations, Dual Fusion Mechanism Prediction model for lncRNA-Disease Association(DFMP-LDA) is proposed based on Graph Attention Network(GAT) and Inductive Martix Completion(IMC).In the first step of DFMP-LDA, a multi-head attention mechanism is introduced to design a GAT with duplex polymerizers, which enhance the features of lncRNA nodes and disease nodes.In the second step, as the traditional GAT cannot be directly applied to the potential lncRNA-disease prediction, IMC technology is introduced to reconstruct the lncRNA-disease association network.The IMC uses the enhanced node features obtained in the first step to improve model accuracy.The results of 5-fold cross-validation show that DFMP-LDA predicts association with an AUC value of 0.932 2 and an AUPR value of 0.770 5, saving 33.89%, 32.17%, 16.12% in time cost compared with DMF-LDA, SDLDA, and TPGLDA, respectively.The experimental results therefore show that DFMP-LDA has better prediction performance than previous prediction frameworks.

Key words: Graph Attention Network(GAT), Inductive Matrix Completion(IMC), association prediction, duplex polymerizer, feature enhancement

中图分类号:

TP391

张奕, 郑婧, 蔡钢生, 王真梅. 基于GAT双聚合运算与归纳式矩阵补全的关联预测[J]. 计算机工程, 2022, 48(12): 72-78.

ZHANG Yi, ZHENG Jing, CAI Gangsheng, WANG Zhenmei. Association Prediction Based on Duplex Polymerize Operation in GAT and Inductive Matrix Completion[J]. Computer Engineering, 2022, 48(12): 72-78.

https://www.ecice06.com/CN/Y2022/V48/I12/72

图/表 6

20230112182659

20230112182704

20230112182707

20230112182710

20230112182714

20230112182718

参考文献

[1] 夏天, 肖丙秀, 郭俊明.长链非编码RNA的作用机制及其研究方法[J].遗传, 2013, 35(3):269-280. XIA T, XIAO B X, GUO J M.Acting mechanisms and research methods of long noncoding RNAs[J].Hereditas, 2013, 35(3):269-280.(in Chinese)
[2] 王利然, 杨丽红, 宁文华, 等.长链非编码RNA调控血管新生的研究进展[J].中国比较医学杂志, 2021, 31(4):143-149. WANG L R, YANG L H, NING W H, et al.Research progress of long non-coding RNA involvement in angiogenesis[J].Chinese Journal of Comparative Medicine, 2021, 31(4):143-149.(in Chinese)
[3] SUN J, SHI H B, WANG Z Z, et al.Inferring novel lncRNA-disease associations based on a random walk model of a lncRNA functional similarity network[J].Molecular BioSystems, 2014, 10(8):2074-2081.
[4] ZHOU M, WANG X J, LI J W, et al.Prioritizing candidate disease-related long non-coding RNAs by walking on the heterogeneous lncRNA and disease network[J].Molecular BioSystems, 2015, 11(3):760-769.
[5] CHEN X.KATZLDA:KATZ measure for the lncRNA-disease association prediction[J].Scientific Reports, 2015, 5(1):16840.
[6] CHEN X, YAN G Y.Novel human lncRNA-disease association inference based on lncRNA expression profiles[J].Bioinformatics, 2013, 29(20):2617-2624.
[7] ZHAO T, XU J, LIU L, et al.Identification of cancer-related lncRNAs through integrating genome, regulome and transcriptome features[J].Molecular BioSystems, 2015, 11(1):126-136.
[8] FU G Y, WANG J, DOMENICONI C, et al.Matrix factorization-based data fusion for the prediction of lncRNA-disease associations[J].Bioinformatics, 2017, 34(9):1529-1537.
[9] 王健宗, 孔令炜, 黄章成, 等.图神经网络综述[J].计算机工程, 2021, 47(4):1-12. WANG J Z, KONG L W, HUANG Z C, et al.Survey of graph neural network[J].Computer Engineering, 2021, 47(4):1-12.(in Chinese)
[10] 金雨澄, 王清钦, 高剑, 等.基于图深度学习的金融文本多标签分类算法[J].计算机工程, 2022, 48(4):16-21. JIN Y C, WANG Q Q, GAO J, et al.Multi-label financial text classification algorithm based on graph deep learning[J].Computer Engineering, 2022, 48(4):16-21.(in Chinese)
[11] 郭嘉琰, 李荣华, 张岩, 等.基于图神经网络的动态网络异常检测算法[J].软件学报, 2020, 31(3):748-762. GUO J Y, LI R H, ZHANG Y, et al.Graph neural network based anomaly detection in dynamic networks[J].Journal of Software, 2020, 31(3):748-762.(in Chinese)
[12] 胡承佐, 王庆梅, 李迪超, 等.基于复杂结构信息的图神经网络序列推荐算法[J].计算机工程, 2022, 48(5):82-90, 97. HU C Z, WANG Q M, LI D C, et al.Sequence recommendation algorithm of graph neural networks based on complex structure information[J].Computer Engineering, 2022, 48(5):82-90, 97.(in Chinese)
[13] 车向北, 康文倩, 邓彬, 等.一种基于图神经网络的SDN路由性能预测模型[J].电子学报, 2021, 49(3):484-491. CHE X B, KANG W Q, DENG B, et al.A prediction model of SDN routing performance based on graph neural network[J].Acta Electronica Sinica, 2021, 49(3):484-491.(in Chinese)
[14] VELICKOVIC P, CUCURULL G, CASANOVA A, et al.Graph attention networks[C]//Proceedings of the 6th International Conference on Learning Representations.Vancouver, Canada:Vancouver Convention Center, 2018:1-12.
[15] VASWANI A, SHAZEER N, PARMAR N, et al.Attention is all you need[EB/OL].[2021-10-21].https://arxiv.org/abs/1706.03762.
[16] LONG Y H, LUO J W, ZHANG Y, et al.Predicting human microbe-disease associations via graph attention networks with inductive matrix completion[J].Briefings in Bioinformatics, 2020, 22(3):1-13.
[17] BIAN C, LEI X J, WU F X.GATCDA:predicting circRNA-disease associations based on graph attention network[J].Cancers, 2021, 13(11):2595.
[18] 顾耀文, 张博文, 郑思, 等.基于图注意力网络的药物ADMET分类预测模型构建方法[J].数据分析与知识发现, 2021, 5(8):76-85. GU Y W, ZHANG B W, ZHENG S, et al.Predicting drug ADMET properties based on graph attention network[J].Data Analysis and Knowledge Discovery, 2021, 5(8):76-85.(in Chinese)
[19] LU C Q, YANG M Y, LUO F, et al.Prediction of lncRNA-disease associations based on inductive matrix completion[J].Bioinformatics, 2018, 34(19):3357-3364.
[20] HUANG L, LI X H, GUO P F, et al.Matrix completion with side information and its applications in predicting the antigenicity of influenza viruses[J].Bioinformatics, 2017, 33(20):3195-3201.
[21] CHEN X, SUN L G, ZHAO Y.NCMCMDA:miRNA-disease association prediction through neighborhood constraint matrix completion[J].Briefings in Bioinformatics, 2020, 22(1):485-496.
[22] KINGMA D P, BA J L.Adam:a method for stochastic optimization[C]//Proceedings of International Conference on Learning Representations.New York, USA:ACM Press, 2015:1-15.
[23] WANG D, WANG J, LU M, et al.Inferring the human microRNA functional similarity and functional network based on microRNA-associated diseases[J].Bioinformatics, 2010, 26(13):1644-1650.
[24] VAN LAARHOVEN T, NABUURS S B, MARCHIORI E.Gaussian interaction profile kernels for predicting drug-target interaction[J].Bioinformatics, 2011, 27(21):3036-3043.
[25] 陈蕾, 陈松灿.矩阵补全模型及其算法研究综述[J].软件学报, 2017, 28(6):1547-1564. CHEN L, CHEN S C.Survey on matrix completion models and algorithms[J].Journal of Software, 2017, 28(6):1547-1564.(in Chinese)
[26] CHEN G, WANG Z Y, WANG D Q, et al.LncRNADisease:a database for long-non-coding RNA-associated diseases[J].Nucleic Acids Research, 2012, 41(D1):D983-D986.
[27] ZENG M, LU C Q, ZHANG F H, et al.SDLDA:lncRNA-disease association prediction based on singular value decomposition and deep learning[J].Methods, 2020, 179:73-80.
[28] ZENG M, LU C, FEI Z, et al.DMFLDA:a deep learning framework for predicting lncRNA-disease associations[J].IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2021, 18(6):2353-2363.
[29] DING L, WANG M H, SUN D D, et al.TPGLDA:novel prediction of associations between lncRNAs and diseases via lncRNA-disease-gene tripartite graph[J].Scientific Reports, 2018, 8(1):1065.

选择文件类型/文献管理软件名称

选择包含的内容