{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,26]],"date-time":"2026-10-26T19:45:27Z","timestamp":1793043927574,"version":"4.5.0"},"reference-count":47,"publisher":"Public Library of Science (PLoS)","issue":"3","license":[{"start":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T00:00:00Z","timestamp":1773878400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>Accurate classification of disease subtypes is a fundamental requirement of precision medicine especially for complex and heterogeneous conditions such as breast cancer and Alzheimer\u2019s disease. Recent advances in graph-based deep learning have shown strong potential in multi-omics integration by modeling inter-sample relationships through similarity networks. Yet, the question of how best to construct these networks remains an open and underexplored challenge. In this work, we present a systematic evaluation of six distinct similarity network construction strategies including Cosine Similarity, Cosine Distance, RBF-based measures, and two hybrid combinations leveraging a graph convolutional network (GCN) integrated with a view correlation discovery network (VCDN) framework for multi-omics disease classification. Using two benchmark datasets (BRCA and ROSMAP), we assessed the impact of each method on classification performance, variance across runs, and statistical robustness. Surprisingly, our results demonstrate that Cosine Similarity outperforms all other metrics, consistently achieving the highest accuracy, F1-score, and AUC, while also showing the lowest standard deviation across cross-validation splits. Despite the growing popularity of kernel-based and hybrid similarity designs, our findings highlight the unique effectiveness of simple angular similarity in capturing biologically meaningful structure in high-dimensional omics data. In our study, we showed that simple yet biologically meaningful similarity measures like Cosine Similarity can outperform more complex techniques in accuracy, consistency, and clarity. This insight sets the stage for building more effective and interpretable graph-based models to support precision medicine.<\/jats:p>","DOI":"10.1371\/journal.pone.0344754","type":"journal-article","created":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T17:41:42Z","timestamp":1773942102000},"page":"e0344754","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":3,"title":["Effects of similarity networks in graph-based multi-omics classification"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0009-0005-6740-1380","authenticated-orcid":true,"given":"Masrafe Bin Hannan","family":"Siam","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Md Rayhan","family":"Khan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0839-5946","authenticated-orcid":true,"given":"Md Fazla","family":"Elahe","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Md Shohel","family":"Arman","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Swarna","family":"Akter","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,3,19]]},"reference":[{"issue":"1","key":"pone.0344754.ref001","doi-asserted-by":"crossref","first-page":"17","DOI":"10.1016\/j.cell.2013.03.002","article-title":"Lessons from the cancer genome","volume":"153","author":"LA Garraway","year":"2013","journal-title":"Cell"},{"issue":"1","key":"pone.0344754.ref002","doi-asserted-by":"crossref","first-page":"30","DOI":"10.1093\/bioinformatics\/btl543","article-title":"Improved breast cancer prognosis through the combination of clinical and genetic markers","volume":"23","author":"Y Sun","year":"2007","journal-title":"Bioinformatics"},{"key":"pone.0344754.ref003","doi-asserted-by":"crossref","first-page":"166","DOI":"10.3389\/fgene.2019.00166","article-title":"SALMON: Survival Analysis Learning With Multi-Omics Neural Networks on Breast Cancer","volume":"10","author":"Z Huang","year":"2019","journal-title":"Front Genet"},{"issue":"6","key":"pone.0344754.ref004","doi-asserted-by":"crossref","first-page":"1248","DOI":"10.1158\/1078-0432.CCR-17-0853","article-title":"Deep Learning-Based Multi-Omics Integration Robustly Predicts Survival in Liver Cancer","volume":"24","author":"K Chaudhary","year":"2018","journal-title":"Clin Cancer Res"},{"issue":"3","key":"pone.0344754.ref005","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pcbi.1011814","article-title":"PathIntegrate: Multivariate modelling approaches for pathway-based multi-omics data integration","volume":"20","author":"C Wieder","year":"2024","journal-title":"PLoS Comput Biol"},{"issue":"3","key":"pone.0344754.ref006","article-title":"Attentive multi-omics integration via graph attention networks for cancer subtype prediction","volume":"25","author":"R Pan","year":"2024","journal-title":"Brief Bioinform"},{"issue":"2","key":"pone.0344754.ref007","first-page":"108172","article-title":"HyperGraphOmics: hypergraph neural networks for multi-omics integration in cancer prognosis","volume":"27","author":"L Zhang","year":"2024","journal-title":"iScience"},{"issue":"6","key":"pone.0344754.ref008","article-title":"Subtype-MGTP: a cancer subtype identification framework based on multi-omics translation","volume":"40","author":"M Xie","year":"2024","journal-title":"Bioinformatics"},{"issue":"12","key":"pone.0344754.ref009","article-title":"Cancer molecular subtyping using limited multi-omics data with missingness","volume":"20","author":"Y Bu","year":"2024","journal-title":"PLoS Comput Biol"},{"issue":"5","key":"pone.0344754.ref010","doi-asserted-by":"crossref","first-page":"463","DOI":"10.1093\/bfgp\/elad013","article-title":"Attention-based GCN integrates multi-omics data for breast cancer subtype classification and patient-specific gene marker identification","volume":"22","author":"H Guo","year":"2023","journal-title":"Brief Funct Genomics"},{"key":"pone.0344754.ref011","doi-asserted-by":"crossref","first-page":"806842","DOI":"10.3389\/fgene.2022.806842","article-title":"MoGCN: A Multi-Omics Integration Method Based on Graph Convolutional Network for Cancer Subtype Analysis","volume":"13","author":"X Li","year":"2022","journal-title":"Front Genet"},{"issue":"1","key":"pone.0344754.ref012","article-title":"GraphFusion: Multi-omics graph fusion network for disease classification and biomarker discovery","volume":"39","author":"W Zhou","year":"2023","journal-title":"Bioinformatics"},{"issue":"1","key":"pone.0344754.ref013","doi-asserted-by":"crossref","first-page":"65","DOI":"10.3390\/genes13010065","article-title":"Identifying Cancer Subtypes Using a Residual Graph Convolution Model on a Sample Similarity Network","volume":"13","author":"W Dai","year":"2021","journal-title":"Genes (Basel)"},{"issue":"5","key":"pone.0344754.ref014","doi-asserted-by":"crossref","first-page":"2788","DOI":"10.3390\/ijms25052788","article-title":"MOGAT: A Multi-Omics Integration Framework Using Graph Attention Networks for Cancer Subtype Prediction","volume":"25","author":"RB Tanvir","year":"2024","journal-title":"Int J Mol Sci"},{"issue":"1","key":"pone.0344754.ref015","doi-asserted-by":"crossref","first-page":"3445","DOI":"10.1038\/s41467-021-23774-w","article-title":"MOGONET integrates multi-omics data using graph convolutional networks allowing patient classification and biomarker identification","volume":"12","author":"T Wang","year":"2021","journal-title":"Nat Commun"},{"issue":"1","key":"pone.0344754.ref016","first-page":"10023","article-title":"NetOmics: A modular network-based approach for multi-omics integration and disease classification","volume":"14","author":"Z Chen","year":"2023","journal-title":"Nat Commun"},{"key":"pone.0344754.ref017","doi-asserted-by":"crossref","first-page":"109","DOI":"10.1515\/9783111504667-008","volume-title":"Drug Discovery and Telemedicine","author":"A Yaqoob","year":"2025"},{"key":"pone.0344754.ref018","doi-asserted-by":"crossref","first-page":"127","DOI":"10.1515\/9783111504667-009","volume-title":"Drug Discovery and Telemedicine","author":"A Yaqoob","year":"2025"},{"issue":"4","key":"pone.0344754.ref019","doi-asserted-by":"crossref","first-page":"3631","DOI":"10.1007\/s10115-025-02340-6","article-title":"Feature selection using game Shapley improved grey wolf optimizer for optimizing cancer classification","volume":"67","author":"S Afreen","year":"2025","journal-title":"Knowl Inf Syst"},{"key":"pone.0344754.ref020","doi-asserted-by":"crossref","first-page":"91","DOI":"10.1515\/9783111504667-007","volume-title":"Drug Discovery and Telemedicine","author":"SU Ahmed","year":"2025"},{"key":"pone.0344754.ref021","doi-asserted-by":"crossref","first-page":"37724","DOI":"10.1109\/ACCESS.2025.3540769","article-title":"Comparative Analysis of Multi-Omics Integration Using Graph Neural Networks for Cancer Classification","volume":"13","author":"F Alharbi","year":"2025","journal-title":"IEEE Access"},{"issue":"2","key":"pone.0344754.ref022","article-title":"RBF gene kernel applications in biological data","volume":"39","author":"X Liu","year":"2023","journal-title":"Bioinformatics"},{"issue":"4","key":"pone.0344754.ref023","article-title":"Predicting drug-disease associations through layer attention graph convolutional network","volume":"22","author":"Z Yu","year":"2021","journal-title":"Brief Bioinform"},{"key":"pone.0344754.ref024","unstructured":"Kipf TN, Welling M. Semi-supervised classification with graph convolutional networks. arXiv preprint. 2016. https:\/\/arxiv.org\/abs\/1609.02907"},{"issue":"4","key":"pone.0344754.ref025","doi-asserted-by":"crossref","first-page":"18","DOI":"10.1109\/MSP.2017.2693418","article-title":"Geometric Deep Learning: Going beyond Euclidean data","volume":"34","author":"MM Bronstein","year":"2017","journal-title":"IEEE Signal Process Mag"},{"issue":"1","key":"pone.0344754.ref026","article-title":"MoBNN: a multi-omics integration framework for cancer subtype classification with graph neural networks and tensor fusion","volume":"23","author":"Q Wen","year":"2022","journal-title":"Brief Bioinform"},{"key":"pone.0344754.ref027","unstructured":"Li A, et al. Multiomics data integration using GCNs. PMC. 2023."},{"issue":"2","key":"pone.0344754.ref028","article-title":"FusionGNN: Interpretable and scalable multi-omics integration using graph neural networks","volume":"25","author":"T Liu","year":"2024","journal-title":"Brief Bioinform"},{"issue":"11","key":"pone.0344754.ref029","first-page":"5654","article-title":"MOGONET-X: Improving multi-omics integration via cross-view interaction modeling and attentional fusion","volume":"26","author":"L Xu","year":"2022","journal-title":"IEEE J Biomed Health Inform"},{"issue":"6","key":"pone.0344754.ref030","article-title":"A multi-omics deep learning model for predicting prognosis and therapeutic responses in breast cancer","volume":"22","author":"W Chen","year":"2021","journal-title":"Brief Bioinform"},{"issue":"1","key":"pone.0344754.ref031","first-page":"1122","article-title":"COSMIx: A cosine similarity-based multi-omics integration framework for disease subtype discovery","volume":"14","author":"M Li","year":"2023","journal-title":"Nature Commun"},{"issue":"5","key":"pone.0344754.ref032","first-page":"2494","article-title":"RBF-GCN: Graph convolutional network with radial basis function for disease subtype classification","volume":"19","author":"R Xu","year":"2020","journal-title":"IEEE\/ACM Trans Computat Biol Bioinform"},{"key":"pone.0344754.ref033","article-title":"Integrative multi-omics analysis in aging and Alzheimer\u2019s disease: the ROSMAP cohort","volume":"22","author":"Y Zhu","year":"2022","journal-title":"Aging Cell"},{"issue":"4","key":"pone.0344754.ref034","first-page":"1234","article-title":"Angular measures for multi-omics similarity networks","volume":"37","author":"J Smith","year":"2021","journal-title":"Bioinformatics"},{"issue":"1","key":"pone.0344754.ref035","first-page":"456","article-title":"Angular-based distances in high-dimensional omics data clustering","volume":"23","author":"R Patel","year":"2022","journal-title":"BMC Genomics"},{"issue":"20","key":"pone.0344754.ref036","first-page":"3508","article-title":"SCSim: single-cell similarity graph construction using cosine and mutual information fusion","volume":"37","author":"M Xu","year":"2021","journal-title":"Bioinformatics"},{"key":"pone.0344754.ref037","volume-title":"Scattered Data Approximation with Radial Basis Functions","author":"H Wendland","year":"2005"},{"key":"pone.0344754.ref038","volume-title":"Learning with kernels: support vector machines, regularization, optimization, and beyond","author":"B Sch\u00f6lkopf","year":"2002"},{"issue":"1","key":"pone.0344754.ref039","first-page":"1","article-title":"CIMLR: multi-omic cancer integration using Gaussian kernels","volume":"6","author":"D Ramazzotti","year":"2018","journal-title":"Cell Syst"},{"issue":"19","key":"pone.0344754.ref040","article-title":"NEMO: integrating multi-omics data using RBF-based similarity networks for clustering","volume":"47","author":"B Wang","year":"2019","journal-title":"Nucleic Acids Res"},{"issue":"4","key":"pone.0344754.ref041","first-page":"456","article-title":"Multi-omics profiling of Alzheimer\u2019s disease in ROSMAP","volume":"3","author":"B Ng","year":"2023","journal-title":"Nat Aging"},{"issue":"6","key":"pone.0344754.ref042","article-title":"The importance of normalization for high-throughput RNA sequencing: a case study","volume":"14","author":"S Hicks","year":"2018","journal-title":"PLoS Computat Biol"},{"issue":"6","key":"pone.0344754.ref043","doi-asserted-by":"crossref","first-page":"520","DOI":"10.1093\/bioinformatics\/17.6.520","article-title":"Missing value estimation methods for DNA microarrays","volume":"17","author":"O Troyanskaya","year":"2001","journal-title":"Bioinformatics"},{"issue":"6","key":"pone.0344754.ref044","article-title":"Hybrid similarity-based graph neural networks for multi-omics data integration in cancer subtype classification","volume":"39","author":"H Song","year":"2023","journal-title":"Bioinformatics"},{"issue":"1","key":"pone.0344754.ref045","article-title":"CosGCN: cosine similarity-based graph convolutional network for patient stratification using multi-omics data","volume":"39","author":"D Kim","year":"2023","journal-title":"Bioinformatics"},{"key":"pone.0344754.ref046","first-page":"1249","article-title":"Multi-omics data integration for subtype classification of breast cancer using deep learning","volume":"19","author":"P Jiang","year":"2021","journal-title":"Computat Struct Biotechnol J"},{"issue":"12","key":"pone.0344754.ref047","first-page":"3572","article-title":"Graph similarity learning for multi-modal biomedical data integration","volume":"36","author":"Y Liu","year":"2020","journal-title":"Bioinformatics"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0344754","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T17:41:53Z","timestamp":1773942113000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0344754"}},"subtitle":[],"editor":[{"given":"Tao","family":"Huang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,3,19]]},"references-count":47,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,3,19]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0344754","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,19]]}}}