CoChat
Sign Up
Cited by 8Machine Learning in BioinformaticsProtein Structure and DynamicsBioinformatics and Genomic Networks

The DOI matches its Crossref record; 7 of the 7 references it cites were found in a registry or an index.

A scalable equivariant graph network framework for precise protein function prediction

Open access
Ran, Zixu, Guo, Xudong, Pan, Tong, Bi, Yue, Hao, Yi, Sun, Heyun, Song, Jiangning, Li, Fuyi·2025·Genome Biology
DOI 10.1186/s13059-025-03886-y

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

BACKGROUND: Protein function research helps in understanding the complex biological processes that occur within cells. However, the intricate nature of protein structures and functions, along with the rapid growth of protein sequence data, presents a pressing challenge to develop efficient computational methods for accurate protein annotation. RESULTS: In this study, we propose ENGINE, a multi-channel deep learning framework designed for robust protein function prediction. ENGINE integrates an equivariant graph convolutional network model to capture geometric features from protein 3D structures, leverages the large language model ESM-C to encode evolutionary and sequence-derived information, and combines an innovative 3D sequence representation that unifies spatial and sequential signals. We demonstrate that ENGINE consistently surpasses current state-of-the-art methods across diverse protein function prediction benchmarks, demonstrating robust generalisation and high predictive accuracy. Beyond performance, ENGINE provides interpretable insights into key sequence features and structural motifs, enabling the identification of functionally critical residues and substructures within proteins. This facilitates a deeper mechanistic understanding of protein function annotation outcomes and supports hypothesis generation for downstream biological studies. CONCLUSION: By offering reliable predictions with biological interpretability, ENGINE contributes to advancing research into cellular processes and disease mechanisms. The model is available at GitHub ( https://github.com/ABILiLab/ENGINE ) and Zenodo ( https://doi.org/10.5281/zenodo.17221153 ), serving as a valuable tool for the broader scientific community.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1186/s13059-025-03886-y

Cites

7 of 7 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Flagged first, then by citations

Research works printed but not linked: 86

  1. Alberts B. The cell as a collection of protein machines: preparing the next generation of molecular biologists. Cell. 1998;92(3):291–4. · DOI · Paper page
  2. Anfinsen CB. Principles that govern the folding of protein chains. Science. 1973;181(4096):223–30. · DOI · Paper page
  3. Pan T, et al. SCREEN: a graph-based contrastive learning tool to infer catalytic residues and assess enzyme mutations. Genomics Proteomics Bioinformatics. 2025;22(6):qzae094. https://doi.org/10.1093/gpbjnl/qzae094. · DOI
  4. Radivojac P, et al. A large-scale evaluation of computational protein function prediction. Nat Methods. 2013;10(3):221–7. · DOI · Paper page
  5. Paananen J, Fortino V. An omics perspective on drug target discovery platforms. Brief Bioinform. 2020;21(6):1937–53. · DOI
  6. Zhang C, Freddolino L, Zhang Y. COFACTOR: improved protein function prediction by combining structure, sequence and protein–protein interaction information. Nucleic Acids Res. 2017;45(W1):W291-9. · DOI · Paper page
  7. Ashburner M, et al. Gene ontology: tool for the unification of biology. Nat Genet. 2000;25(1):25–9. · DOI · Paper page
  8. Kulmanov M, et al. Protein function prediction as approximate semantic entailment. Nat Mach Intell. 2024;6(2):220–8. · DOI · Paper page
  9. Zhang F, et al. Deepfunc: a deep learning framework for accurate prediction of protein functions from protein sequences and interactions. Proteomics. 2019;19(12):1900019. · DOI
  10. Consortium, U. UniProt: a worldwide hub of protein knowledge. Nucleic Acids Res. 2019;47(D1):D506-15. · DOI · Paper page
  11. Torres M, et al. Protein function prediction for newly sequenced organisms. Nat Mach Intell. 2021;3(12):1050–60. · DOI
  12. Eisenberg D, et al. Protein function in the post-genomic era. Nature. 2000;405(6788):823–6. · DOI
  13. Billeter M. Comparison of protein structures determined by NMR in solution and by X-ray diffraction in single crystals. Q Rev Biophys. 1992;25(3):325–77. · DOI
  14. Eisenstein M. The field that came in from the cold. Nat Methods. 2016;13(1):19–23. · DOI
  15. Das S, et al. Functional classification of CATH superfamilies: a domain-based approach for protein function annotation. Bioinformatics. 2015;31(21):3460–7. · DOI
  16. You R, et al. Golabeler: improving sequence-based large-scale protein function prediction by learning to rank. Bioinformatics. 2018;34(14):2465–73. · DOI
  17. You R, et al. NetGO: improving large-scale protein function prediction with massive network information. Nucleic Acids Res. 2019;47(W1):W379–87. · DOI
  18. Cao Y, Shen Y. Tale: transformer-based protein function annotation with joint sequence–label embedding. Bioinformatics. 2021;37(18):2825–33. · DOI
  19. Kulmanov M, Hoehndorf R. DeepGOPlus: improved protein function prediction from sequence. Bioinformatics. 2020;36(2):422–9. · DOI
  20. Pan T, et al. PfresGO: an attention mechanism-based deep-learning approach for protein annotation by integrating gene ontology inter-relationships. Bioinformatics. 2023;39(3):btad094. · DOI
  21. Meng L, Wang X. TAWFN: a deep learning framework for protein function prediction. Bioinformatics. 2024;40(10):btae571. · DOI
  22. Kulmanov M, Hoehndorf R. DeepGOZero: improving protein function prediction from sequence and zero-shot learning based on ontology axioms. Bioinformatics. 2022;38(Supplement_1):i238–45. · DOI
  23. Xia W, et al. PFmulDL: a novel strategy enabling multi-class and multi-label protein function annotation by integrating diverse deep learning methods. Comput Biol Med. 2022;145:105465. · DOI
  24. Berman HM. The protein data bank. Nucleic Acids Res. 2000;28(1):235–42. · DOI · Paper page
  25. Zhang Y, Skolnick J. TM-align: a protein structure alignment algorithm based on the TM-score. Nucleic Acids Res. 2005;33(7):2302–9. · DOI · Paper page
  26. Kadlec J, et al. The postfusion structure of baculovirus gp64 supports a unified view of viral fusion machines. Nat Struct Mol Biol. 2008;15(10):1024–30. · DOI
  27. Rathore S, et al. Cryo-EM structure of the yeast respiratory supercomplex. Nat Struct Mol Biol. 2019;26(1):50–7. · DOI
  28. Fischer ES, et al. The molecular basis of CRL4DDB2/CSA ubiquitin ligase architecture, targeting, and activation. Cell. 2011;147(5):1024–39. · DOI · Paper page
  29. Ying Z, et al. Gnnexplainer: generating explanations for graph neural networks. Adv Neural Inf Process Syst. 2019;32. https://proceedings.neurips.cc/paper_files/paper/2019/hash/d80b7040b773199015de6d3b4293c8ff-Abstract.html.
  30. Papaleo E, et al. The role of protein loops and linkers in conformational dynamics and allostery. Chem Rev. 2016;116(11):6391–423. · DOI
  31. Shehu A, Kavraki LE. Modeling structures and motions of loops in protein molecules. Entropy. 2012;14(2):252–90. · DOI
  32. Kozome D, Sljoka A, Laurino P. Remote loop evolution reveals a complex biological function for chitinase enzymes beyond the active site. Nat Commun. 2024;15(1):3227. · DOI
  33. Gunasekaran K, Nussinov R. Modulating functional loop movements: the role of highly conserved residues in the correlated loop motions. Chembiochem. 2004;5(2):224–30. · DOI
  34. Marcelino AMC, Gierasch LM. Roles of β-turns in protein folding: from peptide models to protein engineering. Biopolymers. 2008;89(5):380–91. · DOI
  35. Fu H, et al. Increasing protein stability by improving beta-turns. Proteins Struct Funct Bioinform. 2009;77(3):491–8. · DOI
  36. Ananthanarayanan V, et al. Structural and functional importance of the β-turn in proteins. Studies on proline-containing peptides. J Biosci. 1985;8:209–21. · DOI
  37. Errington N, Iqbalsyah T, Doig AJ. Structure and stability of the α-Helix: lessons for design. Protein Design Methods Appl. 2006:340:3–26. · DOI
  38. Hol WG. The role of the α-helix dipole in protein function and structure. Prog Biophys Mol Biol. 1985;45(3):149–95. · DOI
  39. Murray KA, et al. Extended β-strands contribute to reversible amyloid formation. ACS Nano. 2022;16(2):2154–63. · DOI
  40. Freire F, et al. Impact of strand length on the stability of parallel-β-sheet secondary structure. Angew Chem Int Ed Engl. 2011;50(37):8735. · DOI
  41. Daffner C, Chelvanayagam G, Argos P. Structural characteristics and stabilizing principles of bent β-strands in protein tertiary architectures. Protein Sci. 1994;3(6):876–82. · DOI
  42. Enkhbayar P, et al. 310-helices in proteins are parahelices. Proteins Struct Funct Bioinform. 2006;64(3):691–9. · DOI
  43. Cooley RB, Arp DJ, Karplus PA. Evolutionary origin of a secondary structure: π-helices as cryptic but widespread insertional variations of α-helices that enhance protein functionality. J Mol Biol. 2010;404(2):232–46. · DOI
  44. Fodje M, Al-Karadaghi S. Occurrence, conformational features and amino acid propensities for the π-helix. Protein Eng. 2002;15(5):353–8. · DOI
  45. Ponting CP, Russell RR. The natural history of protein domains. Annu Rev Biophys Biomol Struct. 2002;31(1):45–71. · DOI
  46. Wilson CA, Kreychman J, Gerstein M. Assessing annotation transfer for genomics: quantifying the relations between protein sequence, structure and function through traditional and probabilistic scores. J Mol Biol. 2000;297(1):233–49. · DOI
  47. Hvidsten TR, et al. A comprehensive analysis of the structure-function relationship in proteins based on local structure similarity. PLoS One. 2009;4(7):e6266. · DOI
  48. Burley SK, et al. RCSB protein data bank: biological macromolecular structures enabling research and education in fundamental biology, biomedicine, biotechnology and energy. Nucleic Acids Res. 2019;47(D1):D464–74. · DOI
  49. Lees JG, et al. Gene3d: multi-domain annotations for protein sequence and comparative genome analysis. Nucleic Acids Res. 2014;42(D1):D240–5. · DOI
  50. Ramon-Maiques S, et al. Structure of acetylglutamate kinase, a key enzyme for arginine biosynthesis and a prototype for the amino acid kinase enzyme family, during catalysis. Structure. 2002;10(3):329–42. · DOI
  51. Massey V. The chemical and biological versatility of riboflavin. Biochem Soc Trans. 2000;28(4):283–96. · DOI
  52. Susini L, et al. TCTP protects from apoptotic cell death by antagonizing bax function. Cell Death Differ. 2008;15(8):1211–20. · DOI
  53. Raj I, Al Hosseini HS, Dioguardi E, et al. Structural basis of egg coat-sperm recognition at fertilization. Cell. 2017;169(7):1315-1326. e17. · DOI
  54. Licht A, et al. Structural and functional characterization of a maltose/maltodextrin ABC transporter comprising a single solute binding domain (MalE) fused to the transmembrane subunit MalF. Res Microbiol. 2019;170(1):1–12. · DOI
  55. Todd AE, Orengo CA, Thornton JM. Evolution of function in protein superfamilies, from a structural perspective. J Mol Biol. 2001;307(4):1113–43. · DOI
  56. Pioszak AA, et al. Molecular recognition of corticotropin-releasing factor by its G-protein-coupled receptor CRFR1. J Biol Chem. 2008;283(47):32900–12. · DOI
  57. Holden HM, et al. Galactokinase: structure, function and role in type II galactosemia. Cell Mol Life Sci. 2004;61(19–20):2471–84. · DOI
  58. Timson DJ, Reece RJ. Functional analysis of disease-causing mutations in human galactokinase. Eur J Biochem. 2003;270(8):1767–74. · DOI
  59. Lichtarge O, Bourne HR, Cohen FE. An evolutionary trace method defines binding surfaces common to protein families. J Mol Biol. 1996;257(2):342–58. · DOI
  60. Yang J, Roy A, Zhang Y. BioLiP: a semi-manually curated database for biologically relevant ligand–protein interactions. Nucleic Acids Res. 2012;41(D1):D1096-103. · DOI
  61. Kursula I, et al. The importance of the conserved Arg191-Asp227 salt bridge of triosephosphate isomerase for folding, stability, and catalysis. FEBS Lett. 2002;518(1–3):39–42. · DOI
  62. Wu S, et al. Diverse roles of assembly factors revealed by structures of late nuclear pre-60S ribosomes. Nature. 2016;534(7605):133–7. · DOI
  63. Li N, et al. Structure of the origin recognition complex bound to DNA replication origin. Nature. 2018;559(7713):217–22. · DOI
  64. Sanghai ZA, et al. Modular assembly of the nucleolar pre-60S ribosomal subunit. Nature. 2018;556(7699):126–9. · DOI
  65. Ohnuma T, et al. Crystal structure and chitin oligosaccharide-binding mode of a “loopful” family GH19 chitinase from rye, Secale cereale, seeds. FEBS J. 2012;279(19):3639–51. · DOI
  66. Taira T, et al. Localization, accumulation, and antifungal activity of chitinases in rye (Secale cereale) seed. Biosci Biotechnol Biochem. 2001;65(12):2710–8. · DOI
  67. Addepalli B, Limbach PA. Pseudouridine in the anticodon of Escherichia coli tRNATyr (QΨA) is catalyzed by the dual specificity enzyme RluF. J Biol Chem. 2016;291(42):22327–37. · DOI
  68. Buchanan SK, et al. Crystal structure of the outer membrane active transporter FepA from Escherichia coli. Nat Struct Biol. 1999;6(1):56–63. · DOI
  69. Parsons JF, et al. Enzymes harboring unnatural amino acids: mechanistic and structural analysis of the enhanced catalytic activity of a glutathione transferase containing 5-fluorotryptophan. Biochemistry. 1998;37(18):6286–94. · DOI
  70. Fu L, et al. CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics. 2012;28(23):3150–2. · DOI · Paper page
  71. Touw WG, et al. A series of PDB-related databanks for everyday needs. Nucleic Acids Res. 2015;43(D1):D364–8. · DOI
  72. Zhou B, et al. Protein engineering with lightweight graph denoising neural networks. J Chem Inf Model. 2024;64(9):3650–61. · DOI
  73. Ganea OE, et al. Independent se (3)-equivariant models for end-to-end rigid protein docking. arXiv preprint arXiv:2111.07786. 2021. https://doi.org/10.48550/arXiv.2111.07786. · DOI
  74. Satorras VG, Hoogeboom E, Welling M. E (n) equivariant graph neural networks. In International conference on machine learning. PMLR; 2021;139:9323–32. https://proceedings.mlr.press/v139/satorras21a.html.
  75. Gainza P, et al. Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning. Nat Methods. 2020;17(2):184–92. · DOI
  76. Ingraham J, et al. Generative models for graph-based protein design. Adv Neural Inf Process Syst. 2019;1417:15820–31. https://doi.org/10.5555/3454287.3455704. · DOI
  77. Hayes T, et al. Simulating 500 million years of evolution with a language model. Science. 2025. https://doi.org/10.1126/science.ads0018. · DOI · Paper page
  78. Kingma DP, Ba J. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980. 2014. https://doi.org/10.48550/arXiv.1412.6980. · DOI · Paper page
  79. Li F, et al. Digerati–a multipath parallel hybrid deep learning framework for the identification of mycobacterial PE/PPE proteins. Comput Biol Med. 2023;163:107155. · DOI
  80. Li F, et al. Prosperousplus: a one-stop and comprehensive platform for accurate protease-specific substrate cleavage prediction and machine-learning model construction. Brief Bioinform. 2023;24(6):bbad372. · DOI
  81. Ran Z, et al. Characterizing secretion system effector proteins with structure-aware graph neural networks and pre-trained language models. IEEE J Biomed Health Inform. 2024. https://doi.org/10.1109/JBHI.2024.3413146. · DOI
  82. Nitish S. Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res. 2014;15:1.
  83. Bi Y, et al. Clarion is a multi-label problem transformation method for identifying mRNA subcellular localizations. Brief Bioinform. 2022;23(6):bbac467. · DOI
  84. Li F, et al. Advancing mRNA subcellular localization prediction with graph neural network and RNA structure. Bioinformatics. 2024;40(8):btae504. · DOI
  85. Wu J, et al. HiFun: homology independent protein function prediction by a novel protein-language self-attention model. Brief Bioinform. 2023;24(5):bbad311. · DOI
  86. Ran Z, Guo X, Pan T, Bi Y, Hao Y, Sun H, Song J, Li F. A scalable equivariant graph network framework for precise protein function prediction. Github. 2025. https://github.com/ABILiLab/ENGINE. · DOI

Metadata

Authors
Ran, Z., Guo, X., Pan, T., Bi, Y., Hao, Y., Sun, H., Song, J., & Li, F.
Affiliations
  • Northwest A&F University (CN)
  • The University of Adelaide (AU)
  • Monash University (AU)
Corresponding author
  • Song, Jiangning
  • Li, Fuyi
ORCID
Venue
Genome Biology
Publisher
Springer Science and Business Media LLC
Published
2025-11-29
Volume
26
Issue
1
Pages
407-407
ISSN
  • 1465-6906 (ISSN-L)
  • 1474-760X
DOI
10.1186/s13059-025-03886-y
OpenAlex
W4416821059
Cited by
8
References
7
Citations by year
2026: 6 · 2025: 2
FWCI
2.40
Type
journal article
Language
en
Open access
Licence
Locations
  • Genome biology (publishedVersion, cc-by-nc-nd)
  • Research Square (acceptedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Natural Science Foundation of China (62202388)
  • National Key Research and Development Program of China (2022YFF1000100)
  • Qin Chuangyuan Innovation and Entrepreneurship Talent Project (QCYRCXM-2022–230)
  • Australia National Health and Medical Research Council (2041439)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
Keywords
protein function prediction, protein 3D structure, protein annotation, interpretable models, structural motifs
MeSH terms
Deep Learning, Protein Conformation, Proteins, Software, Computational Biology, Molecular Sequence Annotation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review