基于异质图注意力网络的miRNA与疾病关联预测算法

doi:10.12263/DZXB.20201116

电子学报 ›› 2022, Vol. 50 ›› Issue (6): 1428-1435.DOI: 10.12263/DZXB.20201116

基于异质图注意力网络的miRNA与疾病关联预测算法

李政伟¹^,², 李佳树¹^,², 尤著宏³, 聂茹², 赵欢², 钟堂波²

^1.中国矿业大学矿山数字化教育部工程研究中心，江苏徐州 221116
^2.中国矿业大学计算机科学与技术学院，江苏徐州 221116
^3.西北工业大学计算机学院，陕西西安 710129

收稿日期:2020-10-12 修回日期:2021-04-08 出版日期:2022-06-25
- 通讯作者:
- 聂茹
- 作者简介:
- 李政伟男，1977年生，河南许昌人.博士.现为中国矿业大学计算机科学与技术学院副教授、硕士生导师.主要研究方向为机器学习与模式识别、生物信息学、计算机视觉等.E-mail： zwli@cumt.edu.cn
  李佳树男，1998年生，江苏徐州人.现为中国矿业大学计算机科学与技术学院研究生.主要研究方向为图神经网络、miRNA与疾病的关联预测等.E-mail： lijiashu7646@163.com
  尤著宏男，1980年生，甘肃兰州人.博士.现为西北工业大学计算机学院教授，博士生导师.主要研究方向为大数据分析、数据挖掘及在生物信息学上的应用等.E-mail： zhuhongyou@nwpu.edu.cn
  聂茹（通讯作者）女，1976年生，江苏徐州人.博士.现为中国矿业大学计算机科学与技术学院副教授，硕士生导师.主要研究方向为生物信息学、机器学习和图像处理等. E-mail：nr@cumt.edu.cn
- 基金资助:
- 国家自然科学基金 (61732012); 科技创新2030—“新一代人工智能”重大项目 (2018AAA0100100)

Associations Prediction Algorithm of MiRNAs and Diseases Based on Heterogeneous Graph Attention Network

LI Zheng-wei¹^,², LI Jia-shu¹^,², YOU Zhu-hong³, NIE Ru², ZHAO Huan², ZHONG Tang-bo²

^1.Engineering Research Center of Mine Digitalization of Ministry of Education, China University of Mining and Technology, Xuzhou, Jiangsu 221116, China
^2.School of Computer Science and Technology, China University of Mining and Technology, Xuzhou, Jiangsu 221116, China
^3.School of Computer Science, Northwestern Polytechnical University, Xi'an, Shaanxi 710129, China

Received:2020-10-12 Revised:2021-04-08 Online:2022-06-25 Published:2022-06-25
- Corresponding author:
- NIE Ru
- Supported by:
- National Natural Science Foundation of China (61732012); National Science and Technology Innovation 2030 -- Major program of New generation of artificial intelligence (2018AAA0100100)

摘要/Abstract

摘要：

众多实验表明，microRNA（miRNA）的异常表达与人类复杂疾病的产生和演化有关.识别miRNA与疾病间的关联有助于促进临床医学的发展.然而，传统的实验方法往往耗时耗力、效率低下，因此迫切需要高效的计算方法对miRNA与疾病间的潜在关联进行预测.本文提出了一种基于异质图注意力网络的端到端的计算模型来预测miRNA与疾病的关联.该方法通过多头注意力机制捕获异质邻居的结构和属性信息，并将其与中心顶点的属性信息进行融合，从而构建出更具表达能力的miRNA和疾病的特征嵌入，进而通过全连接层对miRNA与疾病间的潜在关联进行预测.5折交叉验证结果显示，该模型分别在HMDD v2.0和HMDD v3.0数据集上取得了93.52%和94.82%的AUC值.此外，关于食管肿瘤的病例研究结果显示，该模型预测的前50个miRNA中有48个得到了证实.上述实验结果表明，该模型可作为一种可靠的工具预测候选疾病的相关miRNA.

关键词: 图注意力网络, miRNA-疾病关联, 异质图, 深度学习

Abstract:

Lots of experiments have shown that the abnormal expression of microRNA(miRNA) is related to the evolution and progression of human complex diseases. Identifying associations between miRNAs and diseases is beneficial to promote the development of clinical medicine. However, traditional experimental methods are often time-consuming and inefficient, so there is an urgent need for efficient computational methods to predict the potential associations between miRNAs and diseases. In this paper, we propose an end-to-end computational model based on heterogeneous graph attention network to predict the associations between miRNAs and diseases. This model captures the structure and attribute information of heterogeneous neighbors via the multi-head attention mechanism, and fuses them with the attribute information of the central vertex to generate more representative feature embeddings of miRNAs and diseases, and then predicts the potential associations between miRNAs and diseases through a fully connected layer. The 5-fold cross-validation results show that our model achieves 93.52% and 94.82% AUC values based on HMDD v2.0 and HMDD v3.0 datasets, respectively. In addition, the case study on esophageal neoplasms shows that 48 of the top 50 miRNAs predicted by our model are confirmed. The above experimental results indicate that our model can be used as a reliable tool to predict candidate disease-related miRNAs.

Key words: graph attention network, microRNA-disease associations, heterogeneous graph, deep learning

中图分类号:

TP399

李政伟, 李佳树, 尤著宏, 聂茹, 赵欢, 钟堂波. 基于异质图注意力网络的miRNA与疾病关联预测算法[J]. 电子学报, 2022, 50(6): 1428-1435.

LI Zheng-wei, LI Jia-shu, YOU Zhu-hong, NIE Ru, ZHAO Huan, ZHONG Tang-bo. Associations Prediction Algorithm of MiRNAs and Diseases Based on Heterogeneous Graph Attention Network[J]. Acta Electronica Sinica, 2022, 50(6): 1428-1435.

图/表 8

参考文献 27

1	AMBROSV. The functions of animal microRNAs[J]. Nature, 2004, 431(7006): 350-355.
2	XUP Z, GUOM, HAYB A. MicroRNAs and the regulation of cell death[J]. Trends Genet, 2004, 20(12): 617-624.
3	MELTZERP S. Cancer genomics: Small RNAs with big impacts[J]. Nature, 2005, 435(7043): 745-746.
4	MONSÁLVEZV, MONTES-MORENOS, ARTIGAM J, et al. MicroRNAs as prognostic markers in indolent primary cutaneous B-cell lymphoma[J]. Mod Pathol, 2013, 26(2): 171-181.
5	VÁRALLYAYE, BURGYÁNJ, HAVELDAZ. MicroRNA detection by northern blotting using locked nucleic acid probes[J]. Nat Protoc, 2008, 3(2): 190-196.
6	FREEMANW M, WALKERS J, VRANAK E. Quantitative RT-PCR: Pitfalls and potential[J]. Biotechniques, 1999, 26(1): 112-122, 124-115.
7	BASKERVILLES, BARTELD P. Microarray profiling of microRNAs reveals frequent coexpression with neighboring miRNAs and host genes[J]. Rna, 2005, 11(3): 241-247.
8	SHENZ, ZHANGY H, HANK, et al. MiRNA-disease association prediction with collaborative matrix factorization[J]. Complexity, 2017, 2017: 1-9.
9	JIB Y, YOUZ H, CHENGL, et al. Predicting miRNA-disease association from heterogeneous information network with GraRep embedding model[J]. Scientific Reports, 2020, 10(1): 6658.
10	ZHANGQ H, ZHUL, HUANGD S. High-order convolutional neural network architecture for predicting DNA-protein binding sites[J]. IEEE/ACM Trans Comput Biol Bioinform, 2019, 16(4): 1184-1192.
11	XUW X, ZHUL, HUANGD S. DCDE: An efficient deep convolutional divergence encoding method for human promoter recognition[J]. IEEE Trans Nanobioscience, 2019, 18(2): 136-145.
12	LIY, QIUC X, TUJ, et al. HMDD v2.0: A database for experimentally supported human microRNA and disease associations[J]. Nucleic Acids Research, 2013, 42(D1): D1070-D1074.
13	WANGD, WANGJ, LUM, et al. Inferring the human microRNA functional similarity and functional network based on microRNA-associated diseases[J]. Bioinformatics, 2010, 26(13): 1644-1650.
14	XUANP, HANK, GUOM Z, et al. Prediction of microRNAs associated with human diseases based on weighted k most similar neighbors[J]. PloS one, 2013, 8(8): e70204.
15	VANL T, NABUURSS B, MARCHIORIE. Gaussian interaction profile kernels for predicting drug-target interaction[J]. Bioinformatics, 2011, 27(21): 3036-3043.
16	VASWANIA, SHAZEERN, PARMARN, et al. Attention is all you need[EB/OL]. (2017-06-12)[2020-10-12]..
17	WANGX, JIH Y, SHIC, et al. Heterogeneous graph attention network [EB/OL]. (2019-03-18)[2020-10-12]. .
18	WANGM J, YUL F, ZHENGD, et al. Deep graph library: Towards efficient and scalable deep learning on graphs[EB/OL]. (2017-09-03)[2020-10-12]..
19	CHENX, YANC G, ZHANGX, et al. WBSMDA: Within and between score for miRNA-disease association prediction[J]. Sci Rep, 2016, 6: 21106.
20	CHENX, XIED, WANGL, et al. BNPMDA: Bipartite network projection for miRNA-disease association prediction[J]. Bioinformatics, 2018, 34(18): 3178-3186.
21	CHENX, LIS X, YINJ, et al. Potential miRNA-disease association prediction based on kernelized Bayesian matrix factorization[J]. Genomics, 2020, 112(1): 809-819.
22	XIEG B, FANZ L, SUNY P, et al. WBNPMD: Weighted bipartite network projection for microRNA-disease association prediction[J]. Journal of Translational Medicine, 2019, 17(1): 322.
23	ZHANGL, LIUB L, LIZ W, et al. Predicting miRNA-disease associations by multiple meta-paths fusion graph embedding model[J]. BMC Bioinformatics, 2020, 21(1): 470.
24	MAY J, HET T, GEL X, et al. MiRNA-disease interaction prediction based on kernel neighborhood similarity and multi-network bidirectional propagation[J]. BMC Med Genomics, 2019, 12(10): 185.
25	YUS P, LIANGC, XIAOQ, et al. MCLPMDA: A novel method for miRNA-disease association prediction based on matrix completion and label propagation[J]. Journal of Cellular and Molecular Medicine, 2019, 23(2): 1427-1438.
26	YANGZ, RENF, LIUC N, et al. DbDEMC: A database of differentially expressed miRNAs in human cancers[J]. BMC Genomics, 2010, 11 (4): S5-S5.
27	JIANGQ H, WANGY D, HAOY Y, et al. MiR2Disease: A manually curated database for microRNA deregulation in human disease[J]. Nucleic Acids Research, 2009, 37: D98-D104.

数据集	HMDD v2.0	HMDD v3.0
疾病数目	383	850
miRNA数目	495	1 057
关联数目	5 430	32 226

数据集	HMDD v2.0	HMDD v3.0
疾病数目	383	850
miRNA数目	495	1 057
关联数目	5 430	32 226

测试集	准确率/%	精确率/%	召回率/%	F1值/%
1	85.75	84.08	87.94	85.97
2	87.06	85.94	88.34	87.12
3	86.65	88.50	84.27	86.33
4	85.45	85.69	85.84	85.77
5	85.77	86.30	84.86	85.57
均值	86.14	86.10	86.25	86.15

测试集	准确率/%	精确率/%	召回率/%	F1值/%
1	85.75	84.08	87.94	85.97
2	87.06	85.94	88.34	87.12
3	86.65	88.50	84.27	86.33
4	85.45	85.69	85.84	85.77
5	85.77	86.30	84.86	85.57
均值	86.14	86.10	86.25	86.15

测试集	准确率/%	精确率/%	召回率/%	F1值/%
1	88.44	89.08	87.91	88.49
2	87.31	87.11	87.38	87.24
3	88.04	87.83	88.39	88.11
4	87.66	88.41	86.74	87.57
5	87.82	87.68	87.76	87.72
均值	87.85	88.02	87.64	87.83

基于异质图注意力网络的miRNA与疾病关联预测算法

Associations Prediction Algorithm of MiRNAs and Diseases Based on Heterogeneous Graph Attention Network

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 27

相关文章 15

编辑推荐

Metrics

模型	准确率/%	精确率/%	召回率/%	F1值/%	AUC值/%	AP值/%
中心顶点特征	84.72	85.41	83.77	84.57	92.47	92.15
邻居聚合特征	78.63	71.62	95.03	81.65	91.94	92.01
HGATMDA	86.14	86.10	86.25	86.15	93.52	93.15

模型	AUC值/%
WBSMDA	81.85
BNPMDA	89.80
KBMFMDA	90.08
WBNPMD	91.73
M2GMDA	91.82
KNMBP	93.13
MCLPMDA	93.20
HGATMDA-K1	93.24
HGATMDA-K2	93.34
HGATMDA-K4	93.50
HGATMDA-K8	93.52
HGATMDA-K16	93.49

miRNA (1-25)	证据	miRNA (26-50)	证据
hsa-mir-17	dbDEMC	hsa-mir-181b	dbDEMC
hsa-mir-221	dbDEMC	hsa-mir-106a	dbDEMC
hsa-mir-29a	dbDEMC	hsa-mir-30a	dbDEMC
hsa-mir-16	dbDEMC	hsa-mir-93	dbDEMC
hsa-mir-222	dbDEMC	hsa-mir-23a	dbDEMC
hsa-mir-18a	dbDEMC	hsa-mir-132	dbDEMC
hsa-mir-19b	dbDEMC	hsa-let-7f	dbDEMC
hsa-mir-125b	dbDEMC	hsa-let-7g	dbDEMC
hsa-mir-1	dbDEMC	hsa-mir-23b	dbDEMC
hsa-mir-29b	dbDEMC	hsa-mir-7	dbDEMC
hsa-mir-122	未证实	hsa-mir-133b	dbDEMC
hsa-mir-206	dbDEMC	hsa-mir-125a	dbDEMC
hsa-mir-106b	dbDEMC	hsa-mir-107	dbDEMC, miR2Disease
hsa-mir-200b	dbDEMC	hsa-mir-429	dbDEMC
hsa-let-7i	dbDEMC	hsa-mir-124	dbDEMC
hsa-mir-181a	dbDEMC	hsa-mir-27b	dbDEMC
hsa-let-7e	dbDEMC	hsa-mir-10b	dbDEMC
hsa-mir-146b	dbDEMC	hsa-mir-199b	dbDEMC
hsa-mir-182	dbDEMC	hsa-mir-30c	dbDEMC
hsa-let-7d	dbDEMC	hsa-mir-96	dbDEMC
hsa-mir-142	dbDEMC	hsa-mir-224	dbDEMC
hsa-mir-15b	dbDEMC	hsa-mir-20b	dbDEMC
hsa-mir-24	dbDEMC	hsa-mir-218	dbDEMC
hsa-mir-195	dbDEMC	hsa-mir-103a	未证实
hsa-mir-9	dbDEMC	hsa-mir-26b	dbDEMC

[1]	吴靖, 叶晓晶, 黄峰, 陈丽琼, 王志锋, 刘文犀. 基于深度学习的单帧图像超分辨率重建综述[J]. 电子学报, 2022, 50(9): 2265-2294.
[2]	李雪莹, 王田路, 梁鹏, 王翀. 基于系统模型的用户评论中非功能需求的自动分类[J]. 电子学报, 2022, 50(9): 2079-2089.
[3]	琚长瑞, 秦晓燕, 袁广林, 李豪, 朱虹. 尺度敏感损失与特征融合的快速小目标检测方法[J]. 电子学报, 2022, 50(9): 2119-2126.
[4]	张志昌, 于沛霖, 庞雅丽, 朱林, 曾扬扬. SMGN：用于对话状态跟踪的状态记忆图网络[J]. 电子学报, 2022, 50(8): 1851-1858.
[5]	张亚洲, 俞洋, 朱少林, 陈锐, 戎璐, 梁辉. 一种量子概率启发的对话讽刺识别网络模型[J]. 电子学报, 2022, 50(8): 1885-1893.
[6]	王飞扬, 冀鹏欣, 孙笠, 危倩, 李根, 张忠宝. 一种基于深度学习的动态社交网络用户对齐方法[J]. 电子学报, 2022, 50(8): 1925-1936.
[7]	徐兴荣, 刘聪, 李婷, 郭娜, 任崇广, 曾庆田. 基于双向准循环神经网络和注意力机制的业务流程剩余时间预测方法[J]. 电子学报, 2022, 50(8): 1975-1984.
[8]	欧阳与点, 谢鲲, 谢高岗, 文吉刚. 面向大规模网络测量的数据恢复算法：基于关联学习的张量填充[J]. 电子学报, 2022, 50(7): 1653-1663.
[9]	裴炤, 邱文涛, 王淼, 马苗, 张艳宁. 基于Transformer动态场景信息生成对抗网络的行人轨迹预测方法[J]. 电子学报, 2022, 50(7): 1537-1547.
[10]	彭闯, 王伦文, 胡炜林. 融合深度特征的电磁频谱异常检测算法[J]. 电子学报, 2022, 50(6): 1359-1369.
[11]	杨伟超, 杜宇, 文伟, 侯舒维, 徐常志, 张建华. 基于多重分形谱智能分析的卫星信号调制识别研究[J]. 电子学报, 2022, 50(6): 1336-1343.
[12]	张波, 陆云杰, 秦东明, 邹国建. 一种卷积自编码深度学习的空气污染多站点联合预测模型[J]. 电子学报, 2022, 50(6): 1410-1427.
[13]	廖勇, 李玉杰. 一种轻量化低复杂度的FDD大规模MIMO系统CSI反馈方法[J]. 电子学报, 2022, 50(5): 1211-1217.
[14]	冀振燕, 韩梦豪, 宋晓军, 冯其波. 面向激光光条图像修复的循环相似度映射网络[J]. 电子学报, 2022, 50(5): 1234-1242.
[15]	潘敏婷, 王韫博, 朱祥明, 高思宇, 龙明盛, 杨小康. 基于无标签视频数据的深度预测学习方法综述[J]. 电子学报, 2022, 50(4): 869-886.