{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,3]],"date-time":"2026-06-03T13:45:22Z","timestamp":1780494322674,"version":"3.54.1"},"update-to":[{"DOI":"10.1371\/journal.pcbi.1011330","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2023,12,19]],"date-time":"2023-12-19T00:00:00Z","timestamp":1702944000000}}],"reference-count":49,"publisher":"Public Library of Science (PLoS)","issue":"12","license":[{"start":{"date-parts":[[2023,12,7]],"date-time":"2023-12-07T00:00:00Z","timestamp":1701907200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100012166","name":"National Key Research and Development Program of China","doi-asserted-by":"publisher","award":["NO.2021YFF1200400"],"award-info":[{"award-number":["NO.2021YFF1200400"]}],"id":[{"id":"10.13039\/501100012166","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Major Program of Shenzhen Bay Laboratory","award":["S201101001"],"award-info":[{"award-number":["S201101001"]}]}],"content-domain":{"domain":["www.ploscompbiol.org"],"crossmark-restriction":false},"short-container-title":["PLoS Comput Biol"],"abstract":"<jats:p>Recent advances in deep learning have significantly improved the ability to infer protein sequences directly from protein structures for the fix-backbone design. The methods have evolved from the early use of multi-layer perceptrons to convolutional neural networks, transformers, and graph neural networks (GNN). However, the conventional approach of constructing K-nearest-neighbors (KNN) graph for GNN has limited the utilization of edge information, which plays a critical role in network performance. Here we introduced SPIN-CGNN based on protein contact maps for nearest neighbors. Together with auxiliary edge updates and selective kernels, we found that SPIN-CGNN provided a comparable performance in refolding ability by AlphaFold2 to the current state-of-the-art techniques but a significant improvement over them in term of sequence recovery, perplexity, deviation from amino-acid compositions of native sequences, conservation of hydrophobic positions, and low complexity regions, according to the test by unseen structures, \u201challucinated\u201d structures and diffusion models. Results suggest that low complexity regions in the sequences designed by deep learning, for generated structures in particular, remain to be improved, when compared to the native sequences.<\/jats:p>","DOI":"10.1371\/journal.pcbi.1011330","type":"journal-article","created":{"date-parts":[[2023,12,7]],"date-time":"2023-12-07T13:47:37Z","timestamp":1701956857000},"page":"e1011330","update-policy":"https:\/\/doi.org\/10.1371\/journal.pcbi.corrections_policy","source":"Crossref","is-referenced-by-count":15,"title":["SPIN-CGNN: Improved fixed backbone protein design with contact map-based graph construction and contact graph neural network"],"prefix":"10.1371","volume":"19","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8902-0355","authenticated-orcid":true,"given":"Xing","family":"Zhang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hongmei","family":"Yin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fei","family":"Ling","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jian","family":"Zhan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9958-5699","authenticated-orcid":true,"given":"Yaoqi","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2023,12,7]]},"reference":[{"issue":"1","key":"pcbi.1011330.ref001","doi-asserted-by":"crossref","first-page":"19","DOI":"10.1002\/cbic.200500235","article-title":"Protein-structure prediction by recombination of fragments","volume":"7","author":"J M. Bujnicki","year":"2006","journal-title":"Chembiochem"},{"key":"pcbi.1011330.ref002","doi-asserted-by":"crossref","first-page":"315","DOI":"10.1146\/annurev-biophys-083012-130315","article-title":"Energy functions in de novo protein design: current challenges and future prospects.","volume":"42","author":"Z Li","year":"2013","journal-title":"Annual review of biophysics"},{"issue":"11","key":"pcbi.1011330.ref003","doi-asserted-by":"crossref","first-page":"681","DOI":"10.1038\/s41580-019-0163-x","article-title":"Advances in protein structure prediction and design","volume":"20","author":"B Kuhlman","year":"2019","journal-title":"Nature Reviews Molecular Cell Biology"},{"issue":"1","key":"pcbi.1011330.ref004","doi-asserted-by":"crossref","first-page":"203","DOI":"10.1016\/j.jmb.2004.12.019","article-title":"Energy functions for protein design: adjustment with protein\u2013protein complex affinities, models for the unfolded state, and negative design of solubility and specificity","volume":"347","author":"N Pokala","year":"2005","journal-title":"Journal of Molecular Biology"},{"issue":"1","key":"pcbi.1011330.ref005","doi-asserted-by":"crossref","first-page":"5330","DOI":"10.1038\/ncomms6330","article-title":"Protein design with a comprehensive statistical energy function and boosted by experimental selection for foldability","volume":"5","author":"P Xiong","year":"2014","journal-title":"Nature Communications"},{"issue":"1","key":"pcbi.1011330.ref006","doi-asserted-by":"crossref","first-page":"136","DOI":"10.1093\/bioinformatics\/btz515","article-title":"Increasing the efficiency and accuracy of the ABACUS protein sequence design method","volume":"36","author":"P Xiong","year":"2020","journal-title":"Bioinformatics"},{"key":"pcbi.1011330.ref007","first-page":"545","volume-title":"Methods in Enzymology","author":"A Leaver-Fay","year":"2011"},{"issue":"1","key":"pcbi.1011330.ref008","doi-asserted-by":"crossref","first-page":"86","DOI":"10.1093\/bioinformatics\/btab598","article-title":"De novo protein design by an energy function based on series expansion in distance and orientation dependence","volume":"38","author":"S Liang","year":"2022","journal-title":"Bioinformatics"},{"issue":"10","key":"pcbi.1011330.ref009","first-page":"2565","article-title":"Direct prediction of profiles of sequences compatible with a protein structure by neural networks with fragment-based local and energy-based nonlocal profiles. Proteins: Structure, Function, and","volume":"82","author":"Z Li","year":"2014","journal-title":"Bioinformatics"},{"issue":"6","key":"pcbi.1011330.ref010","doi-asserted-by":"crossref","first-page":"629","DOI":"10.1002\/prot.25489","article-title":"SPIN2: Predicting sequence profiles from protein structures using deep neural networks.","volume":"86","author":"J O\u2019Connell","year":"2018","journal-title":"Proteins: Structure, Function, and Bioinformatics"},{"issue":"1","key":"pcbi.1011330.ref011","first-page":"1","article-title":"Computational protein design with deep learning neural networks.","volume":"8","author":"J Wang","year":"2018","journal-title":"Scientific Reports"},{"issue":"7873","key":"pcbi.1011330.ref012","doi-asserted-by":"crossref","first-page":"583","DOI":"10.1038\/s41586-021-03819-2","article-title":"Highly accurate protein structure prediction with AlphaFold","volume":"596","author":"J Jumper","year":"2021","journal-title":"Nature"},{"issue":"6557","key":"pcbi.1011330.ref013","doi-asserted-by":"crossref","first-page":"871","DOI":"10.1126\/science.abj8754","article-title":"Accurate prediction of protein structures and interactions using a three-track neural network","volume":"373","author":"M Baek","year":"2021","journal-title":"Science"},{"issue":"4","key":"pcbi.1011330.ref014","doi-asserted-by":"crossref","first-page":"292","DOI":"10.1016\/j.cels.2019.03.006","article-title":"End-to-end differentiable learning of protein structure","volume":"8","author":"M. AlQuraishi","year":"2019","journal-title":"Cell Systems"},{"key":"pcbi.1011330.ref015","article-title":"Learning protein structure with a differentiable simulator","author":"J Ingraham","year":"2019","journal-title":"International Conference on Learning Representations"},{"issue":"1","key":"pcbi.1011330.ref016","doi-asserted-by":"crossref","first-page":"391","DOI":"10.1021\/acs.jcim.9b00438","article-title":"To improve protein sequence profile prediction through image captioning on pairwise residue distance map","volume":"60","author":"S Chen","year":"2019","journal-title":"Journal of Chemical Information and Modeling"},{"issue":"3","key":"pcbi.1011330.ref017","doi-asserted-by":"crossref","first-page":"1245","DOI":"10.1021\/acs.jcim.0c00043","article-title":"DenseCPD: improving the accuracy of neural-network-based computational protein sequence design with DenseNet","volume":"60","author":"Y Qi","year":"2020","journal-title":"Journal of Chemical Information and Modeling"},{"issue":"7","key":"pcbi.1011330.ref018","doi-asserted-by":"crossref","first-page":"819","DOI":"10.1002\/prot.25868","article-title":"ProDCoNN: Protein design using a convolutional neural network.","volume":"88","author":"Y Zhang","year":"2020","journal-title":"Proteins: Structure, Function, and Bioinformatics"},{"key":"pcbi.1011330.ref019","first-page":"32","article-title":"Generative models for graph-based protein design","author":"J Ingraham","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"key":"pcbi.1011330.ref020","first-page":"2022","article-title":"Generative de novo protein design with global context.","volume":"10673","author":"C Tan","journal-title":"arXiv preprint arXiv:2204"},{"key":"pcbi.1011330.ref021","article-title":"Learning from protein structure with geometric vector perceptrons","author":"B Jing","year":"2020","journal-title":"arXiv preprint arXiv:2009.01411"},{"key":"pcbi.1011330.ref022","article-title":"Alphadesign: A graph protein design method and benchmark on alphafolddb.","author":"Z Gao","year":"2022","journal-title":"arXiv preprint arXiv:2202.01079"},{"key":"pcbi.1011330.ref023","first-page":"8946","article-title":"Learning inverse folding from millions of predicted structures","author":"C Hsu","year":"2022","journal-title":"International Conference on Machine Learning"},{"issue":"6615","key":"pcbi.1011330.ref024","doi-asserted-by":"crossref","first-page":"49","DOI":"10.1126\/science.add2187","article-title":"Robust deep learning\u2013based protein sequence design using ProteinMPNN","volume":"378","author":"J Dauparas","year":"2022","journal-title":"Science"},{"key":"pcbi.1011330.ref025","article-title":"PiFold: Toward effective and efficient protein inverse folding.","author":"Z Gao","year":"2022","journal-title":"arXiv preprint arXiv:2209.12643"},{"key":"pcbi.1011330.ref026","author":"Z Zheng","journal-title":"Structure-informed Language Models Are Protein DesignersbioRxiv, 2023: 2023.02. 03.526917"},{"issue":"7","key":"pcbi.1011330.ref027","doi-asserted-by":"crossref","first-page":"451","DOI":"10.1038\/s43588-022-00273-6","article-title":"Rotamer-free protein sequence design based on deep learning and self-consistency","volume":"2","author":"Y Liu","year":"2022","journal-title":"Nature Computational Science"},{"issue":"3","key":"pcbi.1011330.ref028","doi-asserted-by":"crossref","first-page":"btad122","DOI":"10.1093\/bioinformatics\/btad122","article-title":"Accurate and efficient protein sequence design through learning concise local environment of residues","volume":"39","author":"B Huang","year":"2023","journal-title":"Bioinformatics"},{"issue":"7897","key":"pcbi.1011330.ref029","doi-asserted-by":"crossref","first-page":"523","DOI":"10.1038\/s41586-021-04383-5","article-title":"A backbone-centred energy function of neural networks for protein design","volume":"602","author":"B Huang","year":"2022","journal-title":"Nature"},{"issue":"6604","key":"pcbi.1011330.ref030","doi-asserted-by":"crossref","first-page":"387","DOI":"10.1126\/science.abn2100","article-title":"Scaffolding protein functional sites using deep learning","volume":"377","author":"J Wang","year":"2022","journal-title":"Science"},{"issue":"7889","key":"pcbi.1011330.ref031","doi-asserted-by":"crossref","first-page":"547","DOI":"10.1038\/s41586-021-04184-w","article-title":"De novo protein design by deep network hallucination","volume":"600","author":"I Anishchenko","year":"2021","journal-title":"Nature"},{"key":"pcbi.1011330.ref032","article-title":"Protein structure and sequence generation with equivariant denoising diffusion probabilistic models","author":"N Anand","year":"2022","journal-title":"arXiv preprint arXiv:2205.15019"},{"key":"pcbi.1011330.ref033","article-title":"Broadly applicable and accurate protein design by integrating structure prediction networks and diffusion generative models","author":"L Watson J","year":"2022","journal-title":"bioRxiv"},{"key":"pcbi.1011330.ref034","first-page":"1","article-title":"Large language models generate functional protein sequences across diverse families","author":"A Madani","year":"2023","journal-title":"Nature Biotechnology"},{"issue":"3","key":"pcbi.1011330.ref035","first-page":"464","article-title":"Research progress of artificial intelligence in designing protein structures","volume":"4","author":"C Zhihang","year":"2023","journal-title":"Synthetic Biology Journal"},{"key":"pcbi.1011330.ref036","doi-asserted-by":"crossref","first-page":"702","DOI":"10.1002\/prot.20264","article-title":"Scoring function for automated assessment of protein structure template quality","volume":"57","author":"Y. Zhang","year":"2004","journal-title":"Proteins"},{"key":"pcbi.1011330.ref037","article-title":"SE(3) diffusion model with application to protein backbone generation.","author":"J Yim","year":"2023","journal-title":"arXiv preprint arXiv:2302.02277"},{"issue":"6637","key":"pcbi.1011330.ref038","doi-asserted-by":"crossref","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":"Z Lin","year":"2023","journal-title":"Science"},{"key":"pcbi.1011330.ref039","first-page":"30","article-title":"Attention is all you need","author":"A Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"pcbi.1011330.ref040","first-page":"510","article-title":"Selective kernel networks","author":"X Li","year":"2019","journal-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition"},{"key":"pcbi.1011330.ref041","article-title":"Decoupled weight decay regularization","author":"I Loshchilov","year":"2017","journal-title":"arXiv preprint arXiv:1711.05101"},{"key":"pcbi.1011330.ref042","first-page":"369","article-title":"Super-convergence: Very fast training of neural networks using large learning rates. Artificial intelligence and machine learning for multi-domain operations applications","volume":"11006","author":"N Smith L","year":"2019","journal-title":"SPIE"},{"key":"pcbi.1011330.ref043","first-page":"32","article-title":"Pytorch: An imperative style, high-performance deep learning library","author":"A Paszke","year":"2019","journal-title":"Advances in Neural Information Processing Systems"},{"issue":"11","key":"pcbi.1011330.ref044","doi-asserted-by":"crossref","first-page":"1422","DOI":"10.1093\/bioinformatics\/btp163","article-title":"Biopython: freely available Python tools for computational molecular biology and bioinformatics","volume":"25","author":"A Cock P J","year":"2009","journal-title":"Bioinformatics"},{"issue":"11","key":"pcbi.1011330.ref045","doi-asserted-by":"crossref","first-page":"e80635","DOI":"10.1371\/journal.pone.0080635","article-title":"Maximum allowed solvent accessibilites of residues in proteins","volume":"8","author":"Z Tien M","year":"2013","journal-title":"PloS One"},{"issue":"2","key":"pcbi.1011330.ref046","doi-asserted-by":"crossref","first-page":"149","DOI":"10.1016\/0097-8485(93)85006-X","article-title":"Statistics of local complexity in amino acid sequences and sequence databases","volume":"17","author":"C Wootton J","year":"1993","journal-title":"Computers & Chemistry"},{"key":"pcbi.1011330.ref047","unstructured":"The NCBI C++ Toolkit (https:\/\/ncbi.github.io\/cxx-toolkit\/) by the National Center for Biotechnology Information, U.S. National Library of Medicine; Bethesda MD, 20894 USA."},{"issue":"22","key":"pcbi.1011330.ref048","doi-asserted-by":"crossref","first-page":"10915","DOI":"10.1073\/pnas.89.22.10915","article-title":"Amino acid substitution matrices from protein blocks","volume":"89","author":"S Henikoff","year":"1992","journal-title":"Proceedings of the National Academy of Sciences"},{"issue":"S3","key":"pcbi.1011330.ref049","first-page":"22","article-title":"Processing and analysis of CASP3 protein structure predictions. Proteins: Structure, Function, and","volume":"37","author":"A Zemla","year":"1999","journal-title":"Bioinformatics"}],"updated-by":[{"DOI":"10.1371\/journal.pcbi.1011330","type":"new_version","label":"New version","source":"publisher","updated":{"date-parts":[[2023,12,19]],"date-time":"2023-12-19T00:00:00Z","timestamp":1702944000000}}],"container-title":["PLOS Computational Biology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1011330","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,12,19]],"date-time":"2023-12-19T13:39:03Z","timestamp":1702993143000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pcbi.1011330"}},"subtitle":[],"editor":[{"given":"Yang","family":"Zhang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2023,12,7]]},"references-count":49,"journal-issue":{"issue":"12","published-online":{"date-parts":[[2023,12,7]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pcbi.1011330","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2023.07.07.548080","asserted-by":"object"}]},"ISSN":["1553-7358"],"issn-type":[{"value":"1553-7358","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,12,7]]}}}