计算机科学 ›› 2025, Vol. 52 ›› Issue (11A): 241000145-9.doi: 10.11896/jsjkx.241000145
胡昭龙, 胡春玲, 胡瑞捷, 郭龙菊
HU Zhaolong, HU Chunling, HU Ruijie, GUO Longju
摘要： 通过深入研究蛋白质-蛋白质相互作用位点(PPIS),能够揭示生命在分子层面运作的深层原理。然而现有方法鉴定PPIS复杂且耗时,需要更精确的模型进行PPIS预测。尽管基于注意力机制和卷积神经网络(CNN)的深度学习方法在PPIS预测方面取得了进展,但在氨基酸特性表征上仍存在局限。为了有效捕捉蛋白质序列中远距离的依赖关系,并准确地表征氨基酸的特性,提出了一种用于预测蛋白质-蛋白质相互作用位点的多特征混合网络(Multi-feature hybrid networks)——PPIS-MFH,通过结合全局序列特征与局部序列特征对PPIS进行预测。对于局部序列特征,PPIS-MFH模型融合了Vision Transformer(ViT)模块,该模块能够捕获蛋白质序列中的远距离依赖性,并提取局部特征。对于全局序列特征,模型PPIS-MFH通过由文本卷积神经网络(TextCNN)并引入注意力机制的文本循环神经网络(TextRNN-Attention)构成的特征交叉网络,利用双向门控循环单元网络来识别蛋白质序列中氨基酸间的内在联系。在4个数据集上对PPIS-MFH模型进行了评估,将其与8种同类方法进行了比较。实验结果显示在大多数指标上,所提方法优于其他的同类方法。
中图分类号:
| [1]DAS S,CHAKRABARTI S.Classification and prediction of protein-protein interaction interface using machine learning algorithm [J].Scientific Reports,2021,11(1):1761. [2]BUTLAND G,PEREGRÍN-ALVAREZ J M,LI J,et al.Interaction network containing conserved and essential protein complexes in Escherichia coli [J].Nature,2005,433(7025):531-537. [3]LI X,LI W,ZENG M,et al.Network-based methods for predicting essential genes or proteins:a survey [J].Briefings in Bioinformatics,2020,21(2):566-583. [4]DE LAS RIVAS J,FONTANILLO C.Protein-protein interac-tions essentials:key concepts to building and analyzing interactome networks [J].PLoS Computational Biology,2010,6(6):e1000807. [5]BRETTNER L M,MASEL J.Protein stickiness,rather thannumber of functional protein-protein interactions,predicts expression noise and plasticity in yeast [J].BMC Systems Biology,2012,6:1-10. [6]TERENTIEV A A,MOLDOGAZIEVA N T,SHAITAN K V.Dynamic proteomics in modeling of the living cell.Protein-protein interactions [J].Biochemistry(Moscow),2009,74:1586-1607. [7]WODAK S J,VLASBLOM J,TURINSKY A L,et al.Protein-protein interaction networks:the puzzling riches [J].Current Opinion in Structural Biology,2013,23(6):941-953. [8]LI Y,GOLDING G B,ILIE L.DELPHI:accurate deep ensemble model for protein interaction sites prediction [J].Bioinformatics,2021,37(7):896-904. [9]HOU Q,DE GEEST P F G,VRANKEN W F,et al.Seeing thetrees through the forest:sequence-based homo-and heteromeric protein-protein interaction sites prediction using random forest [J].Bioinformatics,2017,33(10):1479-1487. [10]HOU Q,LENSINK M F,HERINGA J,et al.Club-martini:se-lectingfavourable interactions amongst available candidates,a coarse-grained simulation approach to scoring docking decoys [J].PloS One,2016,11(5):e0155251. [11]ZHOU Y,JIANG Y,YANG Y.AGAT-PPIS:A novel protein-protein interaction site predictor based on augmented graph attention network with initial residual and identity mapping [J].Briefings in Bioinformatics,2023,24(3):bbad122. [12]PITRE S,DEHNE F,CHAN A,et al.PIPE:a protein-protein interaction prediction engine based on the re-occurring short polypeptide sequences between known interacting protein pairs [J].BMC Bioinformatics,2006,7:1-15. [13]OFRAN Y,ROST B.Predicted protein-protein interaction sites from local sequence information [J].FEBS Letters,2003,544(1／2／3):236-239. [14]MURAKAMI Y,MIZUGUCHI K.Applying the Naïve Bayesclassifier with kernel density estimation to the prediction of protein-protein interaction sites [J].Bioinformatics,2010,26(15):1841-1848. [15]YOUSEF A,CHARKARI N M.A novel methodbased on new adaptive LVQ neural network for predicting protein-protein interactions from protein sequences [J].Journal of Theoretical Biology,2013,336:231-239. [16]SINGH G,DHOLE K,PAI P P,et al.SPRINGS:prediction of protein-protein interaction sites using artificial neural networks [R].PeerJ PrePrints,2014. [17]WANG B,CHEN P,WANG P,et al.Radial basis function neural network ensemble for predicting protein-protein interaction sites in heterocomplexes [J].Protein and Peptide Letters,2010,17(9):1111-1116. [18]KOIKE A,TAKAGI T.Prediction of protein-protein interaction sites using support vector machines [J].Protein Engineering Design and Selection,2004,17(2):165-173. [19]WANG X,YU B,MA A,et al.Protein-protein interaction sites prediction by ensemble random forests with synthetic minority oversampling technique [J].Bioinformatics,2019,35(14):2395-2402. [20]ZENG M,ZHANG F,WU F X,et al.Protein-protein interaction site prediction through combining local and global features with deep neural networks [J].Bioinformatics,2020,36(4):1114-1120. [21]ZHANG B,LI J,QUAN L,et al.Sequence-based prediction of protein-protein interaction sites by simplified long short-term memory network [J].Neurocomputing,2019,357:86-100. [22]LU S,LI Y,NAN X,et al.Attention-based convolutional neural networks for protein-protein interaction site prediction [C]//2021 IEEE International Conference on Bioinformatics and Biomedicine(BIBM).IEEE,2021:141-144. [23]CONG H,LIU H,CAO Y,et al.Protein-protein interaction site prediction by modelensembling with hybrid feature and self-attention [J].BMC Bioinformatics,2023,24(1):456. [24]WANG X,YU B,MA A,et al.Protein-protein interaction sites prediction by ensemble random forests with synthetic minority oversampling technique [J].Bioinformatics,2019,35(14):2395-2402. [25]JOOSTEN R P,TE BEEK T A H,KRIEGER E,et al.A series of PDB related databases for everyday needs [J].Nucleic Acids Research,2010,39(suppl_1):D411-D419. [26]KABSCH W,SANDER C.Dictionary of protein secondarystructure:pattern recognition of hydrogen-bonded and geometrical features [J].Biopolymers:Original Research on Biomolecules,1983,22(12):2577-2637. [27]WANG J,YANG B,REVOTE J,et al.POSSUM:a bioinformatics toolkit for generating numerical sequence feature descriptors based onPSSM profiles [J].Bioinformatics,2017,33(17):2756-2758. [28]WODAK S J,VLASBLOM J,TURINSKY A L,et al.Protein-protein interaction networks:the puzzling riches[J].Current Opinion in Structural Biology,2013,23(6):941-953. |
|
||
首页