{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,5]],"date-time":"2025-11-05T07:24:50Z","timestamp":1762327490951,"version":"build-2065373602"},"reference-count":62,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,11,5]],"date-time":"2025-11-05T00:00:00Z","timestamp":1762300800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100002842","name":"Chiang Mai University","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100002842","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100004704","name":"National Research Council of Thailand","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100004704","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Bioinform."],"abstract":"<jats:p>\n                    Accurate prediction of antibody paratopes is a critical challenge in structure-limited, high-throughput discovery workflows. We present ParaDeep, a lightweight and interpretable deep learning framework for residue-level paratope prediction directly from amino acid sequences. ParaDeep integrates bidirectional long short-term memory networks with one-dimensional convolutional layers to capture both long-range sequence context and local binding motifs. We systematically evaluated 30 model configurations varying in encoding schemes, convolutional kernel sizes, and antibody chain types. In five-fold cross-validation, heavy (H) chain models achieved the highest performance (F1 = 0.856 \u00b1 0.014, MCC = 0.842 \u00b1 0.015), outperforming light (L) chain models (F1 = 0.774 \u00b1 0.023, MCC = 0.772 \u00b1 0.022). On an independent blind test set, ParaDeep attained F1 = 0.723 and MCC = 0.685 for H chains, and F1 = 0.607 and MCC = 0.587 for L chains, representing a 27% MCC improvement over the sequence-based baseline Parapred. Chain-specific modeling revealed that heavy chains provide stronger sequence-based predictive signals, while light chains benefit more from structural context. ParaDeep approaches the performance of state-of-the-art structure-based methods on heavy chains while requiring only sequence input, enabling faster and broader applicability without the computational cost of 3D modeling. Its efficiency and scalability make it well-suited for early-stage antibody discovery, repertoire profiling, and therapeutic design, particularly in the absence of structural data. The implementation is freely available at\n                    <jats:ext-link>https:\/\/github.com\/PiyachatU\/ParaDeep<\/jats:ext-link>\n                    , with Python (PyTorch) code and a Google Colab interface for ease of use.\n                  <\/jats:p>","DOI":"10.3389\/fbinf.2025.1684042","type":"journal-article","created":{"date-parts":[[2025,11,5]],"date-time":"2025-11-05T06:26:36Z","timestamp":1762323996000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["ParaDeep: sequence-based deep learning for residue-level paratope prediction using chain-aware BiLSTM-CNN models"],"prefix":"10.3389","volume":"5","author":[{"given":"Piyachat","family":"Udomwong","sequence":"first","affiliation":[]},{"given":"Thanathat","family":"Pamonsupornwichit","sequence":"additional","affiliation":[]},{"given":"Kanchanok","family":"Kodchakorn","sequence":"additional","affiliation":[]},{"given":"Chatchai","family":"Tayapiwatana","sequence":"additional","affiliation":[]}],"member":"1965","published-online":{"date-parts":[[2025,11,5]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"3832","DOI":"10.1016\/j.molimm.2008.05.022","article-title":"Analysis and improvements to kabat and structurally correct numbering of antibody variable domains","volume":"45","author":"Abhinandan","year":"2008","journal-title":"Mol. Immunol."},{"key":"B2","doi-asserted-by":"publisher","first-page":"108856","DOI":"10.1016\/j.celrep.2021.108856","article-title":"A compact vocabulary of paratope-epitope interactions enables predictability of antibody-antigen binding","volume":"34","author":"Akbar","year":"2021","journal-title":"Cell Rep."},{"key":"B3","doi-asserted-by":"publisher","first-page":"831","DOI":"10.1038\/nbt.3300","article-title":"Predicting the sequence specificities of DNA- and RNA-Binding proteins by deep learning","volume":"33","author":"Alipanahi","year":"2015","journal-title":"Nat. Biotechnol."},{"key":"B4","doi-asserted-by":"publisher","first-page":"1619","DOI":"10.2741\/2786","article-title":"Humanization of antibodies","volume":"13","author":"Almagro","year":"2008","journal-title":"Front. Biosci."},{"key":"B5","doi-asserted-by":"publisher","first-page":"44","DOI":"10.3390\/antib8030044","article-title":"Phage display libraries for antibody therapeutic discovery and development","volume":"8","author":"Almagro","year":"2019","journal-title":"Antibodies (Basel)."},{"key":"B6","doi-asserted-by":"publisher","first-page":"e21","DOI":"10.23915\/distill.00021","article-title":"Computing receptive fields of convolutional neural networks","volume":"4","author":"Araujo","year":"2019","journal-title":"Distill"},{"key":"B7","doi-asserted-by":"publisher","first-page":"e0141287","DOI":"10.1371\/journal.pone.0141287","article-title":"Continuous distributed representation of biological sequences for deep proteomics and genomics","volume":"10","author":"Asgari","year":"2015","journal-title":"PLOS ONE"},{"key":"B8","doi-asserted-by":"publisher","first-page":"235","DOI":"10.1093\/nar\/28.1.235","article-title":"The protein data bank","volume":"28","author":"Berman","year":"2000","journal-title":"Nucleic Acids Res."},{"key":"B9","doi-asserted-by":"publisher","first-page":"249","DOI":"10.1016\/j.neunet.2018.07.011","article-title":"A systematic study of the class imbalance problem in convolutional neural networks","volume":"106","author":"Buda","year":"2018","journal-title":"Neural Netw."},{"key":"B10","first-page":"28","article-title":"Optimization of the sliding window size for protein structure prediction","author":"Chen","year":"2006"},{"key":"B11","doi-asserted-by":"publisher","first-page":"1596","DOI":"10.1002\/jcc.20918","article-title":"Prediction of protein structural class using novel evolutionary collocation\u2010based sequence representation","volume":"29","author":"Chen","year":"2008","journal-title":"J. Comput. Chem."},{"key":"B12","doi-asserted-by":"publisher","first-page":"btac732","DOI":"10.1093\/bioinformatics\/btac732","article-title":"Paragraph\u2014Antibody paratope prediction using graph neural networks with minimal feature vectors","volume":"39","author":"Chinery","year":"2022","journal-title":"Bioinformatics"},{"volume-title":"Masked language modeling for proteins via linearly scalable long-context transformers","year":"2020","author":"Choromanski","key":"B13"},{"key":"B14","doi-asserted-by":"publisher","first-page":"901","DOI":"10.1016\/0022-2836(87)90412-8","article-title":"Canonical structures for the hypervariable regions of immunoglobulins","volume":"196","author":"Chothia","year":"1987","journal-title":"J. Mol. Biol."},{"key":"B15","doi-asserted-by":"crossref","DOI":"10.1145\/1143844.1143874","article-title":"The relationship between precision-recall and ROC curves","author":"Davis","year":"2006"},{"key":"B16","doi-asserted-by":"publisher","first-page":"298","DOI":"10.1093\/bioinformatics\/btv552","article-title":"ANARCI: antigen receptor numbering and receptor classification","volume":"32","author":"Dunbar","year":"2016","journal-title":"Bioinformatics"},{"key":"B17","doi-asserted-by":"publisher","first-page":"D1140","DOI":"10.1093\/nar\/gkt1043","article-title":"SAbDab: the structural antibody database","volume":"42","author":"Dunbar","year":"2014","journal-title":"Nucleic Acids Res."},{"key":"B18","doi-asserted-by":"publisher","first-page":"7112","DOI":"10.1109\/tpami.2021.3095381","article-title":"ProtTrans: toward understanding the language of life through self-supervised learning","volume":"44","author":"Elnaggar","year":"2022","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell."},{"key":"B19","doi-asserted-by":"publisher","first-page":"487","DOI":"10.1016\/0022-2836(92)91010-m","article-title":"Antibody framework residues affecting the conformation of the hypervariable loops","volume":"224","author":"Foote","year":"1992","journal-title":"J. Mol. Biol."},{"key":"B20","first-page":"1243","article-title":"Convolutional Sequence to Sequence Learning","author":"Gehring","year":"2017"},{"key":"B21","doi-asserted-by":"publisher","first-page":"3168","DOI":"10.1038\/s41467-021-23303-9","article-title":"Structure-based protein function prediction using graph convolutional networks","volume":"12","author":"Gligorijevi\u0107","year":"2021","journal-title":"Nat. Commun."},{"volume-title":"Deep learning","year":"2016","author":"Goodfellow","key":"B22"},{"key":"B23","doi-asserted-by":"publisher","first-page":"2403","DOI":"10.1093\/bioinformatics\/bty1006","article-title":"Improving prediction of protein secondary structure, backbone angles, solvent accessibility and contact numbers by using predicted contact maps and an ensemble of recurrent and residual convolutional neural networks","volume":"35","author":"Hanson","year":"2019","journal-title":"Bioinformatics"},{"key":"B24","doi-asserted-by":"publisher","first-page":"1263","DOI":"10.1109\/tkde.2008.239","article-title":"Learning from imbalanced data","volume":"21","author":"He","year":"2009","journal-title":"IEEE Trans. Knowl. Data Eng."},{"key":"B25","doi-asserted-by":"publisher","first-page":"843","DOI":"10.1093\/bioinformatics\/btv665","article-title":"Highly accurate sequence-based prediction of half-sphere exposures of amino acid residues in proteins","volume":"32","author":"Heffernan","year":"2016","journal-title":"Bioinformatics"},{"key":"B26","doi-asserted-by":"publisher","first-page":"1735","DOI":"10.1162\/neco.1997.9.8.1735","article-title":"Long short-term memory","volume":"9","author":"Hochreiter","year":"1997","journal-title":"Neural Comput"},{"key":"B27","doi-asserted-by":"publisher","first-page":"195","DOI":"10.1006\/jmbi.1999.3091","article-title":"Protein secondary structure prediction based on position-specific scoring matrices11Edited by G. Von Heijne","volume":"292","author":"Jones","year":"1999","journal-title":"J. Mol. Biol."},{"key":"B28","doi-asserted-by":"publisher","first-page":"583","DOI":"10.1038\/s41586-021-03819-2","article-title":"Highly accurate protein structure prediction with AlphaFold","volume":"596","author":"Jumper","year":"2021","journal-title":"Nature"},{"key":"B29","doi-asserted-by":"publisher","first-page":"29141","DOI":"10.1038\/s41598-024-80940-y","article-title":"ParaAntiProt provides paratope prediction using antibody and protein language models","volume":"14","author":"Kalemati","year":"2024","journal-title":"Sci. Rep."},{"key":"B30","doi-asserted-by":"crossref","DOI":"10.3115\/v1\/D14-1181","volume-title":"Convolutional neural networks for sentence classification","author":"Kim","year":"2014"},{"key":"B31","doi-asserted-by":"publisher","first-page":"255","DOI":"10.1038\/nprot.2016.169","article-title":"The ClusPro web server for protein\u2013protein docking","volume":"12","author":"Kozakov","year":"2017","journal-title":"Nat. Protoc."},{"key":"B32","doi-asserted-by":"publisher","first-page":"621","DOI":"10.1093\/protein\/gzt043","article-title":"Antibody i-Patch prediction of the antibody binding site improves rigid local antibody\u2013antigen docking","volume":"26","author":"Krawczyk","year":"2013","journal-title":"Protein Eng. Des. and Sel."},{"key":"B33","doi-asserted-by":"publisher","first-page":"84","DOI":"10.1145\/3065386","article-title":"ImageNet classification with deep convolutional neural networks","volume":"60","author":"Krizhevsky","year":"2017","journal-title":"Commun. ACM"},{"key":"B34","doi-asserted-by":"publisher","first-page":"W521","DOI":"10.1093\/nar\/gks480","article-title":"Paratome: an online tool for systematic identification of antigen-binding regions in antibodies based on sequence or structure","volume":"40","author":"Kunik","year":"2012","journal-title":"Nucleic Acids Res."},{"key":"B35","doi-asserted-by":"publisher","first-page":"507","DOI":"10.1093\/protein\/gzs024","article-title":"Computer-aided antibody design","volume":"25","author":"Kuroda","year":"2012","journal-title":"Protein Eng. Des. and Sel."},{"key":"B36","doi-asserted-by":"publisher","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","author":"LeCun","year":"1998","journal-title":"Proc. IEEE"},{"key":"B37","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1038\/nature14539","article-title":"Deep learning","volume":"521","author":"LeCun","year":"2015","journal-title":"Nature"},{"key":"B38","doi-asserted-by":"publisher","first-page":"100513","DOI":"10.1016\/j.patter.2022.100513","article-title":"Deciphering the language of antibodies using self-supervised learning","volume":"3","author":"Leem","year":"2022","journal-title":"Patterns (N Y)"},{"key":"B39","doi-asserted-by":"publisher","first-page":"2944","DOI":"10.1093\/bioinformatics\/bty305","article-title":"Parapred: antibody paratope prediction using convolutional and recurrent neural networks","volume":"34","author":"Liberis","year":"2018","journal-title":"Bioinformatics"},{"key":"B40","doi-asserted-by":"publisher","first-page":"1123","DOI":"10.1126\/science.ade2574","article-title":"Evolutionary-scale prediction of atomic-level protein structure with a language model","volume":"379","author":"Lin","year":"2023","journal-title":"Science"},{"volume-title":"Efficient estimation of word representations in vector space. International conference on learning representations","year":"2013","author":"Mikolov","key":"B41"},{"key":"B42","doi-asserted-by":"publisher","first-page":"btaf062","DOI":"10.1093\/bioinformatics\/btaf062","article-title":"ParaSurf: a surface-based deep learning approach for paratope\u2013antigen interaction prediction","volume":"41","author":"Papadopoulos","year":"2025","journal-title":"Bioinformatics"},{"key":"B43","article-title":"PyTorch: an imperative style, high-performance deep learning library","volume-title":"Advances in Neural Information Processing Systems (NeurIPS)","author":"Paszke","year":"2019"},{"key":"B44","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/N18-1202","article-title":"Deep Contextualized Word Representations","volume-title":"Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies","author":"Peters","year":"2018"},{"key":"B45","doi-asserted-by":"publisher","first-page":"3996","DOI":"10.1093\/bioinformatics\/btaa263","article-title":"Learning context-aware structural representations to predict antigen and antibody binding interfaces","volume":"36","author":"Pittala","year":"2020","journal-title":"Bioinformatics"},{"key":"B46","doi-asserted-by":"publisher","first-page":"100059","DOI":"10.1074\/mcp.ra120.002314","article-title":"Antibody binding epitope mapping (AbMap) of hundred antibodies in a single run","volume":"20","author":"Qi","year":"2021","journal-title":"Mol. and Cell. Proteomics"},{"key":"B47","doi-asserted-by":"publisher","first-page":"9689","DOI":"10.48550\/arXiv.1906.08230","article-title":"Evaluating protein transfer learning with TAPE","volume":"32","author":"Rao","year":"2019","journal-title":"Adv. Neural Inf. Process. Syst."},{"volume-title":"Deciphering antibody affinity maturation with language models and weakly supervised learning","year":"2021","author":"Ruffolo","key":"B48"},{"key":"B49","doi-asserted-by":"publisher","first-page":"100406","DOI":"10.1016\/j.patter.2021.100406","article-title":"Antibody structure prediction using interpretable deep learning","volume":"3","author":"Ruffolo","year":"2022","journal-title":"Patterns (N Y)"},{"key":"B50","doi-asserted-by":"publisher","first-page":"40","DOI":"10.1002\/prot.21078","article-title":"Prediction of continuous B-cell epitopes in an antigen using recurrent neural network","volume":"65","author":"Saha","year":"2006","journal-title":"Proteins"},{"key":"B51","doi-asserted-by":"publisher","first-page":"2673","DOI":"10.1109\/78.650093","article-title":"Bidirectional recurrent neural networks","volume":"45","author":"Schuster","year":"1997","journal-title":"IEEE Trans. Signal Process."},{"key":"B52","doi-asserted-by":"publisher","first-page":"302","DOI":"10.3389\/fimmu.2013.00302","article-title":"The structural basis of antibody-antigen recognition","volume":"4","author":"Sela-Culang","year":"2013","journal-title":"Front. Immunol."},{"key":"B53","doi-asserted-by":"crossref","DOI":"10.1109\/BigData47090.2019.9005997","article-title":"The performance of LSTM and BiLSTM in forecasting time series","author":"Siami-Namini","year":"2019"},{"key":"B54","doi-asserted-by":"publisher","first-page":"497","DOI":"10.1002\/prot.22309","article-title":"Toward high-resolution homology modeling of antibody Fv regions and application to antibody-antigen docking","volume":"74","author":"Sivasubramanian","year":"2009","journal-title":"Proteins"},{"key":"B55","doi-asserted-by":"publisher","first-page":"3031","DOI":"10.1016\/j.jmb.2015.07.016","article-title":"Updates to the integrated protein-protein interaction benchmarks: docking benchmark version 5 and affinity benchmark version 2","volume":"427","author":"Vreven","year":"2015","journal-title":"J. Mol. Biol."},{"key":"B56","doi-asserted-by":"publisher","first-page":"941","DOI":"10.1016\/j.sbi.2013.08.002","article-title":"Protein\u2013protein interaction networks: the puzzling riches","volume":"23","author":"Wodak","year":"2013","journal-title":"Curr. Opin. Struct. Biol."},{"key":"B57","doi-asserted-by":"publisher","first-page":"752","DOI":"10.1109\/tcbb.2013.75","article-title":"Sequence-based prediction of microRNA-Binding residues in proteins using cost-sensitive laplacian support vector machines","volume":"10","author":"Wu","year":"2013","journal-title":"IEEE\/ACM Trans. Comput. Biol. Bioinforma."},{"key":"B58","first-page":"3780","article-title":"A unified view of multi-label performance measures","volume-title":"Proceedings of the 34th International Conference on Machine Learning (ICML)","author":"Wu","year":"2017"},{"key":"B59","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1016\/j.cbpa.2021.04.004","article-title":"Protein sequence design with deep generative models","volume":"65","author":"Wu","year":"2021","journal-title":"Curr. Opin. Chem. Biol."},{"key":"B60","doi-asserted-by":"publisher","first-page":"37","DOI":"10.1016\/s1074-7613(00)00006-6","article-title":"Diversity in the CDR3 region of VH is sufficient for Most antibody specificities","volume":"13","author":"Xu","year":"2000","journal-title":"Immunity"},{"key":"B61","doi-asserted-by":"publisher","first-page":"i121","DOI":"10.1093\/bioinformatics\/btw255","article-title":"Convolutional neural network architectures for predicting DNA\u2013Protein binding","volume":"32","author":"Zeng","year":"2016","journal-title":"Bioinformatics"},{"key":"B62","doi-asserted-by":"publisher","first-page":"RP104934","DOI":"10.7554\/elife.104934","article-title":"A comprehensive antigen-antibody complex database unlocking insights into interaction interface","volume":"14","author":"Zhou","year":"2025","journal-title":"eLife"}],"container-title":["Frontiers in Bioinformatics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbinf.2025.1684042\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,5]],"date-time":"2025-11-05T06:26:40Z","timestamp":1762324000000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fbinf.2025.1684042\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,5]]},"references-count":62,"alternative-id":["10.3389\/fbinf.2025.1684042"],"URL":"https:\/\/doi.org\/10.3389\/fbinf.2025.1684042","relation":{},"ISSN":["2673-7647"],"issn-type":[{"type":"electronic","value":"2673-7647"}],"subject":[],"published":{"date-parts":[[2025,11,5]]},"article-number":"1684042"}}