基于图神经网络的IL-6诱导肽预测方法

doi:10.12141/j.issn.1000-565X.240242

华南理工大学学报(自然科学版) ›› 2025, Vol. 53 ›› Issue (5): 109-117.doi: 10.12141/j.issn.1000-565X.240242

基于图神经网络的IL-6诱导肽预测方法

曹瑞芬¹, 胡维玲¹, 李强生¹, 宾艳南², 郑春厚³

^1.安徽大学计算机科学与技术学院，安徽合肥 230601
^2.安徽大学物质科学与信息技术研究院，安徽合肥 230601
^3.安徽大学人工智能学院，安徽合肥 230601

收稿日期:2024-05-22 出版日期:2025-05-25 发布日期:2024-11-08
作者简介:曹瑞芬（1981—），女，博士，教授，博士生导师，主要从事人工智能、医学图像处理和生物医学信息处理等研究。E-mail： rfcao@ahu.edu.cn
基金资助:
国家自然科学基金项目(62373001);国家重点研发计划项目(2020YFA0908700);安徽省高校协同创新项目(GXXT-2021-030)

Prediction Method of IL-6 Inducing Peptides Based on Graph Neural Network

CAO Ruifen¹, HU Weiling¹, LI Qiangsheng¹, BIN Yannan², ZHENG Chunhou³

^1.School of Computer Science and Technology，Anhui University，Hefei 230601，Anhui，China
^2.Institutes of Physical Science and Information Technology，Anhui University，Hefei 230601，Anhui，China
^3.School of Artificial Intelligence，Anhui University，Hefei 230601，Anhui，China

Received:2024-05-22 Online:2025-05-25 Published:2024-11-08
About author:曹瑞芬（1981—），女，博士，教授，博士生导师，主要从事人工智能、医学图像处理和生物医学信息处理等研究。E-mail： rfcao@ahu.edu.cn
Supported by:
the National Natural Science Foundation of China(62373001);the National Key Research and Development Program of China(2020YFA0908700)

摘要/Abstract

摘要：

白细胞介素6（简称IL-6）是一种高多效性的糖蛋白因子，可以调节先天性免疫、适应性免疫以及代谢的各个方面，包括糖酵解、脂肪酸氧化和氧化磷酸化等。许多研究已证明，病毒感染患者体内的IL-6表达和释放量显著增加，并且与疾病的严重程度呈正相关，因此，识别IL-6诱导肽并探究其作用机制，对于开发免疫治疗以及疾病严重程度生物标志物具有重要的意义。目前对于IL-6诱导肽的识别大多采用传统的机器学习方法，特征选择与提取较为复杂，且需要依赖领域专家知识。鉴于此，该文提出一种基于图神经网络的IL-6诱导肽预测方法SFGNN-IL6。该方法根据所预测的IL-6诱导肽的结构特征，通过阈值筛选距离信息构建邻接矩阵，结合氨基酸的编码方式（One-hot编码、位置编码和BLOSUM62编码）提取氨基酸节点特征并进行图表示；然后，采用图注意力机制层和图卷积神经网络层，由双通道分别提取多视角特征，既关注节点权重的更新，也考虑节点信息的更新；最后，将两类特征进行融合，用于IL-6诱导肽的分类。实验结果验证了该方法的有效性。

关键词: IL-6诱导肽, 图神经网络, 结构特征, 图注意力机制, 图卷积神经网络

Abstract:

Interleukin-6 (IL-6) is a highly multifunctional glycoprotein factor that can regulate both innate and adaptive immunity as well as various aspects of metabolism, including glycolysis, fatty acid oxidation and oxidative phosphorylation. Many studies have shown that the expression and release of IL-6 in patients infected with viruses significantly increase, and are positively correlated with the severity of the disease. Therefore, identifying IL-6 inducing peptides and exploring their action mechanisms are very important for developing immune therapies and dia-gnostic biomarker for the severity of diseases. Currently, the identification methods of IL-6 inducing peptides mostly use traditional machine learning, in which feature selection and extraction are rather complex, and field expert knowledge are required. In view of this problem, this paper proposes a novel graph neural network-based prediction method of IL-6 inducing peptides named SFGNN-IL6. In this method, the predicted structural characteristics of IL-6 inducing peptides are used to construct the adjacency matrix by screening the distance information according to the threshold, and the node features of amino acids are extracted using One-hot encoding, position encoding and BLOSUM62 encoding, and are then graph-represented. Moreover, graph attention mechanism layers and graph convolutional neural network layers are used as dual channels to separately extract features, considering both the update of node weights and the update of node information. Finally, the two types of features are fused for the classification of IL-6 inducing peptides. Experimental results validate that the proposed method is effective.

Key words: IL-6 inducing peptide, graph neural network, structural feature, graph attention mechanism, graph convolutional neural network

中图分类号:

TP183

曹瑞芬, 胡维玲, 李强生, 宾艳南, 郑春厚. 基于图神经网络的IL-6诱导肽预测方法[J]. 华南理工大学学报(自然科学版), 2025, 53(5): 109-117.

CAO Ruifen, HU Weiling, LI Qiangsheng, BIN Yannan, ZHENG Chunhou. Prediction Method of IL-6 Inducing Peptides Based on Graph Neural Network[J]. Journal of South China University of Technology(Natural Science Edition), 2025, 53(5): 109-117.

图/表 14

图1

表1

图2

图3

表2

图4

表3

图5

图6

图7

图8

表4

表5

表6

参考文献 22

1	PENCE B D ．Aging and monocyte immunometabolism in COVID-19［J］．Aging，2021，13（7）：9154-9155．
2	SÁEZ-CIRIÓN A， SERETI I ．Immunometabolism and HIV-1 pathogenesis：food for thought［J］．Nature Reviews Immunology，2021，21（1）：5-19．
3	LU R， ZHAO X， LI J，et al ．Genomic characterisation and epidemiology of 2019 novel coronavirus： implications for virus origins and receptor binding［J］．The Lancet，2020，395（10224）：565-574．
4	KISHIMOTO T ．IL-6：from its discovery to clinical applications［J］．International Immunology，2010，22（5）：347-352．
5	LI Y-S， REN H-C， CAO J-H ．Roles of interleukin-6-mediated immunometabolic reprogramming in COVID-19 and other viral infection-associated diseases［J］．International Immunopharmacology，2022，110：109005/1-15．
6	OKABAYASHI T， KARIWA H， YOKOTA S I，et al ．Cytokine regulation in SARS coronavirus infection compared to other respiratory virus infections［J］．Journal of Medical Virology，2006，78（4）：417-424．
7	NOTZ Q， SCHMALZING M， WEDEKINK F，et al ．Pro-and anti-inflammatory responses in severe COVID-19-induced acute respiratory distress syndrome—an observational pilot study［J］．Frontiers in Immunology，2020，11：581338/1-13．
8	MANAVALAN B， BASITH S， LEE G ．Comparative analysis of machine learning-based approaches for the identification of therapeutic peptides targeting COVID-19［J］．Briefings in Bioinformatics，2022，23（1）：bbab412/1-16．
9	NAGPAL G， USMANI S S， DHANDA S K，et al ．Computer-aided designing of immunosuppressive peptides based on IL-10 inducing potential［J］．Scientific Reports，2017，7（1）：42851/1-10．
10	DHALL A， PATIYAL S， SHARMA N，et al ．Computer-aided prediction and design of IL-6 inducing peptides： IL-6 plays a crucial role in COVID-19［J］．Briefings in Bioinformatics，2020，22（2）：936-945．
11	CHAROENKWAN P， CHIANGJONG W， NANTASENAMAT C，et al ．StackIL6：a stacking ensemble model for improving the prediction of IL-6 inducing peptides［J］．Briefings in Bioinformatics，2021，22（6）：bbab172/1-13．
12	YAN K， LV H， GUO Y，et al ．sAMPpred-GAT： prediction of antimicrobial peptide by graph attention network and predicted peptide structure［J］．Bioin-formatics，2022，39（1）：btac715/1-8．
13	VASWANI A， SHAZEER N， PARMAR N，et al ．Attention is all you need［M］．GUYONI，LUXBURGU V，BENGIOS，alet，ed．Advances in Neural Information Processing Systems 30．San Diego：NIPS，2017．
14	CHEN Z， ZHAO P， LI F，et al ．iFeature： a python package and web server for features extraction and selection from protein and peptide sequences［J］．Bioinformatics，2018，34（14）：2499-2502．
15	REMMERT M， BIEGERT A， HAUSER A，et al ．HHblits： lightning-fast iterative protein sequence searching by HMM-HMM alignment［J］．Nature Me-thods，2012，9（2）：173-175．
16	DU Z， SU H， WANG W，et al ．The trRosetta server for fast and accurate protein structure prediction［J］．Nature Protocols，2021，16（12）：5634-5651．
17	VELIČKOVIĆ P， CUCURULL G， CASANOVA A，et al ．Graph attention networks［C］∥Proceedings of ICLR 2018．［S.l.］：ICLR，2018：10903/1-12．
18	KIPF T N， WELLING M ．Semi-supervised classification with graph convolutional networks［C］∥Proceedings of ICLR 2017．［S.l.］：ICLR，2017：02907/1-10．
19	RAO B， ZHANG L， ZHANG G ．ACP-GCN：the identification of anticancer peptides based on graph convolution networks［J］．IEEE Access，2020，8：176005-176011．
20	CAO R， HE C， WEI P，et al ．Prediction of circRNA-disease associations based on the combination of multi-head graph attention network and graph convolutional network［J］．Biomolecules，2022，12：932//1-17．
21	DU J， ZHOU Y， LIU P，et al ．Parameter-free loss for class-imbalanced deep learning in image classification［J］．IEEE Transactions on Neural Networks Learning Systems，2021，34（6）：3234-3240．
22	PANG Y， YAO L， XU J，et al ．Integrating transformer and imbalanced multi-label learning to identify antimicrobial peptides and their functional activities［J］．Bioinformatics，2022，38（24）：5368-5374．

数据集	IL-6诱导肽数量/条	非IL-6诱导肽数量/条
训练集	292	2 393
测试集	73	598

阈值/nm	B	S	T	M	A_UC
0.8	0.705±0.014	0.691±0.066	0.718±0.077	0.278±0.035	0.781±0.021
0.9	0.723±0.019	0.637±0.074	0.809±0.096	0.341±0.054	0.799±0.019
1.0	0.706±0.013	0.651±0.087	0.762±0.114	0.301±0.047	0.794±0.013
1.1	0.692±0.040	0.571±0.155	0.813±0.093	0.297±0.046	0.790±0.020

GCN层数	GAT层数	B	S	T	M	A_UC
1	1	0.723±0.026	0.736±0.067	0.709±0.066	0.296±0.038	0.787±0.033
	2	0.723±0.019	0.637±0.074	0.809±0.096	0.341±0.054	0.799±0.019
	3	0.697±0.016	0.624±0.115	0.771±0.142	0.300±0.058	0.788±0.023
2	1	0.709±0.026	0.674±0.069	0.745±0.038	0.286±0.036	0.791±0.026
	2	0.720±0.032	0.718±0.059	0.722±0.093	0.313±0.062	0.794±0.016
	3	0.722±0.020	0.691±0.054	0.752±0.031	0.309±0.029	0.792±0.028

不平衡处理	B	S	T	M	A_UC
正样本随机过采样	0.778	0.643	0.913	0.488	0.868
SMOTE过采样	0.786	0.726	0.846	0.432	0.864
负样本随机欠采样	0.737	0.794	0.680	0.306	0.812
损失函数加权重	0.801	0.753	0.849	0.456	0.875

损失函数	B	S	T	M	A_UC
WCE Loss	0.801	0.753	0.849	0.456	0.875
CE Loss	0.728	0.507	0.948	0.469	0.865
Focal Loss	0.791	0.822	0.761	0.395	0.869
ASL Loss	0.765	0.740	0.791	0.374	0.862

基于图神经网络的IL-6诱导肽预测方法

Prediction Method of IL-6 Inducing Peptides Based on Graph Neural Network

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 14

参考文献 22

相关文章 10

编辑推荐

Metrics

本文评价

方法	B	S	T	M	A_UC
SFGNN-IL6	0.801	0.753	0.849	0.456	0.875
StackIL6	0.795	0.849	0.741	0.393	0.841
IL-6Pred	0.743	0.753	0.732		0.830

[1]	胡广华, 代志刚, 王清辉. 基于图神经网络的B-Rep模型加工特征识别方法[J]. 华南理工大学学报(自然科学版), 2025, 53(5): 20-31.
[2]	蔡晓东, 周青松, 叶青. 基于动态邻域采样的社交推荐模型[J]. 华南理工大学学报(自然科学版), 2024, 52(2): 32-41.
[3]	赵荣超, 吴百礼, 陈祝云, 等. 多尺度时空信息融合驱动的图神经网络故障诊断方法[J]. 华南理工大学学报(自然科学版), 2023, 51(12): 42-52.
[4]	许伦辉, 余佳芯, 裴明阳, 等. 基于几何路网结构和强化学习的车辆重定位策略[J]. 华南理工大学学报(自然科学版), 2023, 51(10): 99-109.
[5]	陆以勤, 潘周双, 张洋, 等. 基于交互式连接图注意力网络的知识图谱补全方法[J]. 华南理工大学学报(自然科学版), 2022, 50(12): 13-19.
[6]	蔡晓东, 曾志杨. AFGSRec：一种自适应融合全局协同特征的社交推荐模型[J]. 华南理工大学学报(自然科学版), 2022, 50(12): 71-79.
[7]	李福海胡丁涵胡志明靳贺松冯志华王奕彬余泳江. 不同因素对碱激发粉煤灰砂浆强度的影响及机理研究[J]. 华南理工大学学报(自然科学版), 2019, 47(5): 139-146.
[8]	张云清余红发王甲春. 气泡结构特征对混凝土抗盐冻性能的影响[J]. 华南理工大学学报（自然科学版）, 2010, 38(11): 7-11.
[9]	高学. 基于动态规划的联机手写汉字分割方法[J]. 华南理工大学学报（自然科学版）, 2007, 35(1): 70-73,79.
[10]	彭晓宏袁铁　彭晓春. 含蒽核心不对称共轭苯乙炔树形聚合物的合成[J]. 华南理工大学学报(自然科学版), 2003, 31(1): 18-21,25.