{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,5]],"date-time":"2026-10-05T17:58:30Z","timestamp":1791223110817,"version":"4.1.0"},"reference-count":20,"publisher":"Public Library of Science (PLoS)","issue":"10","license":[{"start":{"date-parts":[[2026,10,5]],"date-time":"2026-10-05T00:00:00Z","timestamp":1791158400000},"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":"\n Background<\/jats:title>\n Graph attention networks (GATs) are increasingly applied to transcriptomic data because they integrate gene-network structure while producing attention weights that are often interpreted as indicators of biological importance. However, whether attention-derived explanations reliably reflect biologically meaningful signals has received little systematic evaluation, particularly in the small, heterogeneous cohorts common in psychiatric transcriptomics.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n We systematically re-evaluated a GAT using peripheral blood RNA-sequencing data from 76 individuals (39 ADHD, 37 controls), including 16 discordant monozygotic twin pairs. To maximize analytical rigor, we implemented a leakage-aware pipeline, incorporating twin-aware group-stratified cross-validation, fold-wise gene selection, corrected transcript-to-gene mapping, and validation through repeated cross-validation, permutation testing, three complementary feature-importance methods, orthogonal differential-expression and pathway analyses, and data-quality controls.<\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n Across 20 repeated cross-validation splits, GAT achieved a slightly higher mean AUC than a graph convolutional network (mean AUC 0.624 vs. 0.599) and outperformed five classical machine-learning models on a prespecified split. However, GAT performance was not statistically distinguishable from a rigorously matched permutation-derived null distribution (p\u2009=\u20090.327), while an exploratory sample-size calculation indicated that roughly twice the current sample size would be needed to detect an effect this size at conventional power. Independent validation analyses converged on the same conclusion: attention-derived gene rankings were significantly negatively correlated with both SHAP and permutation importance, whereas differential-expression analyses---including discordant twin-pair comparisons---and pathway enrichment identified no reproducible biological signals after multiple-testing correction. Recovery of established co-expression modules and the absence of blood-cell marker differences supported pipeline validity.<\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n Graph-based models may show modest predictive gains, but predictive performance and biological interpretation are distinct scientific questions. Our findings demonstrate that attention-derived importance should be regarded as hypothesis-generating rather than mechanistic evidence unless independently validated, illustrating why rigorous, multi-level validation is essential before biological conclusions are drawn from graph neural networks in transcriptomic research.<\/jats:p>\n <\/jats:sec>","DOI":"10.1371\/journal.pone.0359656","type":"journal-article","created":{"date-parts":[[2026,10,5]],"date-time":"2026-10-05T17:41:26Z","timestamp":1791222086000},"page":"e0359656","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Rigorous validation of graph-based network analysis reappraises biologically coherent ADHD-associated transcriptomic modules in peripheral blood"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8543-6660","authenticated-orcid":true,"given":"Nehal M.","family":"Ali","sequence":"first","affiliation":[{"name":"School of Information Technology, Newgiza University","place":["Cairo, Egypt"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,10,5]]},"reference":[{"issue":"4","key":"pone.0359656.ref001","first-page":"1042","article-title":"ADHD prevalence estimates across three decades","volume":"44","author":"GV Polanczyk","year":"2015","journal-title":"Int J Epidemiol"},{"key":"pone.0359656.ref002","doi-asserted-by":"crossref","first-page":"789","DOI":"10.1016\/j.neubiorev.2021.01.022","article-title":"The World Federation of ADHD international consensus statement","volume":"128","author":"SV Faraone","year":"2021","journal-title":"Neurosci Biobehav Rev"},{"key":"pone.0359656.ref003","first-page":"934943","article-title":"Transcriptomic analysis in ADHD: A systematic review","volume":"13","author":"DT Leffa","year":"2022","journal-title":"Front Psychiatry"},{"key":"pone.0359656.ref004","first-page":"819782","article-title":"Altered inflammatory response in peripheral blood from children with ADHD","volume":"13","author":"HC Chuang","year":"2022","journal-title":"Front Psychiatry"},{"key":"pone.0359656.ref005","doi-asserted-by":"crossref","first-page":"120024","DOI":"10.1109\/ACCESS.2021.3109069","article-title":"A Novel Approach of Transcriptomic microRNA analysis using text mining methods: An early detection of multiple sclerosis disease","volume":"9","author":"NM Ali","year":"2021","journal-title":"IEEE Access"},{"key":"pone.0359656.ref006","doi-asserted-by":"crossref","unstructured":"Ali NM, et al. Early rheumatoid arthritis detection by miRNA data analysis using a hybrid CNN-LSTM deep learning model. IEEE. 2023.","DOI":"10.1109\/IMSA58542.2023.10217733"},{"issue":"10","key":"pone.0359656.ref007","first-page":"288","article-title":"Integrated analysis of microRNA and mRNA in ADHD","volume":"9","author":"N Nuzziello","year":"2019","journal-title":"Brain Sci"},{"issue":"1","key":"pone.0359656.ref008","first-page":"160","article-title":"Biomarker discovery in ADHD: RNA sequencing of whole blood","volume":"13","author":"JT Nigg","year":"2020","journal-title":"BMC Med Genomics"},{"issue":"5","key":"pone.0359656.ref009","first-page":"651","article-title":"Multi-ancestry GWAS of ADHD identifies 27 risk loci","volume":"53","author":"EM Bruxel","year":"2021","journal-title":"Nat Genet"},{"key":"pone.0359656.ref010","volume-title":"Statistical power analysis for the behavioral sciences","author":"J Cohen","year":"1988","edition":"2"},{"key":"pone.0359656.ref011","first-page":"5998","article-title":"Attention is all you need.","volume-title":"Advances in Neural Information Processing Systems 30 (NeurIPS 2017)","author":"A Vaswani","year":"2017"},{"issue":"43","key":"pone.0359656.ref012","doi-asserted-by":"crossref","first-page":"15545","DOI":"10.1073\/pnas.0506580102","article-title":"Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles","volume":"102","author":"A Subramanian","year":"2005","journal-title":"Proc Natl Acad Sci U S A"},{"issue":"6","key":"pone.0359656.ref013","doi-asserted-by":"crossref","first-page":"417","DOI":"10.1016\/j.cels.2015.12.004","article-title":"The Molecular Signatures Database (MSigDB) hallmark gene set collection","volume":"1","author":"A Liberzon","year":"2015","journal-title":"Cell Syst"},{"issue":"2","key":"pone.0359656.ref014","doi-asserted-by":"crossref","first-page":"125","DOI":"10.1038\/s41576-022-00532-2","article-title":"Obtaining genetics insights from deep learning via explainable artificial intelligence","volume":"24","author":"G Novakovsky","year":"2023","journal-title":"Nat Rev Genet"},{"issue":"1","key":"pone.0359656.ref015","article-title":"Graph representation learning in bioinformatics: Trends, methods and applications","volume":"23","author":"H-C Yi","year":"2022","journal-title":"Brief Bioinform"},{"issue":"4","key":"pone.0359656.ref016","doi-asserted-by":"crossref","first-page":"259","DOI":"10.1080\/14737159.2024.2333277","article-title":"A state-of-the-art overview of candidate diagnostic biomarkers for Attention-deficit\/hyperactivity disorder (ADHD)","volume":"24","author":"V Parlatini","year":"2024","journal-title":"Expert Rev Mol Diagn"},{"key":"pone.0359656.ref017","article-title":"Shared genetic architecture between ADHD and intelligence varies across ADHD subtypes","author":"Q Zhao","year":"2026","journal-title":"BMC Med"},{"issue":"2","key":"pone.0359656.ref018","doi-asserted-by":"crossref","first-page":"198","DOI":"10.1038\/s41588-022-01285-8","article-title":"Genome-wide analyses of ADHD identify 27 risk loci, refine the genetic architecture and implicate several cognitive domains","volume":"55","author":"D Demontis","year":"2023","journal-title":"Nat Genet"},{"issue":"6","key":"pone.0359656.ref019","article-title":"A comprehensive review and evaluation of graph neural networks for non-coding RNA and complex disease associations","volume":"24","author":"X Hu","year":"2023","journal-title":"Brief Bioinform"},{"key":"pone.0359656.ref020","doi-asserted-by":"crossref","first-page":"160","DOI":"10.1016\/j.euroneuro.2020.11.005","article-title":"Transcriptome profiling in adult attention-deficit hyperactivity disorder","volume":"41","author":"N Mortimer","year":"2020","journal-title":"Eur Neuropsychopharmacol"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0359656","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,5]],"date-time":"2026-10-05T17:41:33Z","timestamp":1791222093000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0359656"}},"subtitle":[],"editor":[{"given":"Feng","family":"Liu","sequence":"first","affiliation":[{"name":"Tianjin Medical University General Hospital","place":["CHINA"]}],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,10,5]]},"references-count":20,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2026,10,5]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0359656","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,5]]}}}