Artificial intelligence (AI) has revolutionized structural biology by enabling the prediction of protein structures, biomolecular complexes, and interaction interfaces with unprecedented accuracy. The field has evolved from sequence motif searches and conventional machine learning algorithms to deep learning architectures and protein models capable of learning structural and functional representations directly from protein sequences. Recent advances, including protein language models (PLMs), AlphaFold2/3, ESMFold, and Boltz-2, have transformed the prediction of protein–protein interactions by integrating sequence, evolutionary, and structural information while reducing reliance on experimental structural data. These methods have expanded the ability to identify binding interfaces, characterize biomolecular complexes, and investigate interactions involving intrinsically disordered proteins and short linear motifs (SLiMs), although accurately modeling dynamic and context-dependent interactions remains a significant challenge. In this review, we summarize the evolution of AI-driven approaches for predicting protein–protein interactions, from classical machine learning to modern foundation models, highlighting their strengths, limitations, and emerging applications. The 14-3-3 protein family serves as a representative example to illustrate how AI has progressively improved the identification of phosphorylation-dependent binding motifs and interaction partners. Finally, we discuss future perspectives for integrating protein language models, structural prediction, and experimental data to achieve a more comprehensive understanding of protein interaction networks and to accelerate the discovery of regulatory mechanisms and therapeutic targets.
Citation: Sergio Müller, Ignacio Moreno Guerci, Diego Martin Bustos. AI-driven prediction of 14-3-3–intrinsically disordered protein interactions: From SLiM discovery to structural modeling of dynamic binding interfaces[J]. AIMS Biophysics, 2026, 13(4): 424-442. doi: 10.3934/biophy.2026022
Artificial intelligence (AI) has revolutionized structural biology by enabling the prediction of protein structures, biomolecular complexes, and interaction interfaces with unprecedented accuracy. The field has evolved from sequence motif searches and conventional machine learning algorithms to deep learning architectures and protein models capable of learning structural and functional representations directly from protein sequences. Recent advances, including protein language models (PLMs), AlphaFold2/3, ESMFold, and Boltz-2, have transformed the prediction of protein–protein interactions by integrating sequence, evolutionary, and structural information while reducing reliance on experimental structural data. These methods have expanded the ability to identify binding interfaces, characterize biomolecular complexes, and investigate interactions involving intrinsically disordered proteins and short linear motifs (SLiMs), although accurately modeling dynamic and context-dependent interactions remains a significant challenge. In this review, we summarize the evolution of AI-driven approaches for predicting protein–protein interactions, from classical machine learning to modern foundation models, highlighting their strengths, limitations, and emerging applications. The 14-3-3 protein family serves as a representative example to illustrate how AI has progressively improved the identification of phosphorylation-dependent binding motifs and interaction partners. Finally, we discuss future perspectives for integrating protein language models, structural prediction, and experimental data to achieve a more comprehensive understanding of protein interaction networks and to accelerate the discovery of regulatory mechanisms and therapeutic targets.
| [1] |
Ward JJ, Sodhi JS, McGuffin LJ, et al. (2004) Prediction and functional analysis of native disorder in proteins from the three kingdoms of life. J Mol Biol 337: 635-645. https://doi.org/10.1016/j.jmb.2004.02.002 
|
| [2] |
Dyson HJ, Wright PE (2005) Intrinsically unstructured proteins and their functions. Nat Rev Mol Cell Biol 6: 197-208. https://doi.org/10.1038/nrm1589
|
| [3] |
Van Roey K, Uyar B, Weatheritt RJ, et al. (2014) Short linear motifs: Ubiquitous and functionally diverse protein interaction modules directing cell regulation. Chem Rev 114: 6733-6778. https://doi.org/10.1021/cr400585q
|
| [4] |
Tcherkasskaya O, Davidson EA, Uversky VN (2003) Biophysical constraints for protein structure prediction. J Proteome Res 2: 37-42. https://doi.org/10.1021/pr025552q
|
| [5] |
Abramson J, Adler J, Dunger J, et al. (2024) Accurate structure prediction of biomolecular interactions with AlphaFold 3. Nature 630: 493-500. https://doi.org/10.1038/s41586-024-07487-w
|
| [6] |
Lin Z, Akin H, Rao R, et al. (2023) Evolutionary-scale prediction of atomic-level protein structure with a language model. Science 379: 1123-1130. https://doi.org/10.1126/science.ade2574
|
| [7] | Passaro S, Corso G, Wohlwend J, et al. (2025) Boltz-2: Towards accurate and efficient binding affinity prediction. BioRxiv . https://doi.org/10.1101/2025.06.14.659707 |
| [8] |
Omidi A, Møller MH, Malhis N, et al. (2024) AlphaFold-multimer accurately captures interactions and dynamics of intrinsically disordered protein regions. Natl Acad Sci USA 121: e2406407121. https://doi.org/10.1073/pnas.2406407121
|
| [9] |
Akdel M, Pires DEV, Pardo EP, et al. (2022) A structural biology community assessment of AlphaFold2 applications. Nat Struct Mol Biol 29: 1056-1067. https://doi.org/10.1038/s41594-022-00849-w
|
| [10] | Ille AM, Markosian C, Burley SK, et al. (2025) Human protein interactome structure prediction at scale with boltz-2. bioRxiv . https://doi.org/10.1101/2025.07.03.663068 |
| [11] |
Uhart M, Iglesias AA, Bustos DM (2011) Structurally constrained residues outside the binding motif are essential in the interaction of 14-3-3 and phosphorylated partner. J Mol Biol 406: 552-557. https://doi.org/10.1016/j.jmb.2010.12.043
|
| [12] |
Moore BW (1969) Acidic protein. Chemical Architecture of the Nervous System . Boston: Springer 93-99. https://doi.org/10.1007/978-1-4615-7154-4_6
|
| [13] |
Aitken A (2006) 14-3-3 proteins: A historic overview. Semin Cancer Biol 16: 162-172. https://doi.org/10.1016/j.semcancer.2006.03.005
|
| [14] |
Gogl G, Tugaeva KV, Eberling P, et al. (2021) Hierarchized phosphotarget binding by the seven human 14-3-3 isoforms. Nat Commun 12: 1677. https://doi.org/10.1038/s41467-021-21908-8
|
| [15] |
Masone D, Uhart M, Bustos DM (2017) On the role of residue phosphorylation in 14-3-3 partners: AANAT as a case study. Sci Rep 7: 46114. https://doi.org/10.1038/srep46114
|
| [16] |
Gardino AK, Smerdon SJ, Yaffe MB (2006) Structural determinants of 14-3-3 binding specificities and regulation of subcellular localization of 14-3-3-ligand complexes: A comparison of the x-ray crystal structures of all human 14-3-3 isoforms. Semin Cancer Biol 16: 173-182. https://doi.org/10.1016/j.semcancer.2006.03.007
|
| [17] |
Sluchanko NN, Bustos DM (2019) Intrinsic disorder associated with 14-3-3 proteins and their partners. Progress in Molecular Biology and Translational Science 166: 19-61. https://doi.org/10.1016/BS.PMBTS.2019.03.007
|
| [18] |
Muslin AJ, Tanner JW, Allen PM, et al. (1996) Interaction of 14-3-3 with signaling proteins is mediated by the recognition of phosphoserine. Cell 84: 889-897. https://doi.org/10.1016/s0092-8674(00)81067-3
|
| [19] |
Wang B, Yang H, Liu Y-C, et al. (1999) Isolation of high-affinity peptide antagonists of 14-3-3 proteins by phage display. Biochemistry 38: 12499-12504. https://doi.org/10.1021/bi991353h
|
| [20] |
Pozuelo Rubio M, Geraghty KM, Wong BHC, et al. (2004) 14-3-3-affinity purification of over 200 human phosphoproteins reveals new links to regulation of cellular metabolism, proliferation and trafficking. Biochem J 379: 395-408. https://doi.org/10.1042/BJ20031797
|
| [21] |
Uhart M, Bustos DM (2013) Human 14-3-3 paralogs differences uncovered by cross-talk of phosphorylation and lysine acetylation. Plos One 8: e55703. https://doi.org/10.1371/journal.pone.0055703
|
| [22] |
Stevers LM, Sijbesma E, Botta M, et al. (2018) Modulators of 14-3-3 protein–protein interactions. J Med Chem 61: 3755-3778. https://doi.org/10.1021/acs.jmedchem.7b00574
|
| [23] |
Wright PE, Dyson HJ (2014) Intrinsically disordered proteins in cellular signalling and regulation. Nat Rev Mol Cell Biol 16: 18-29. https://doi.org/10.1038/nrm3920
|
| [24] |
Dinkel H, Van Roey K, Michael S, et al. (2014) ELM 2014–a database of eukaryotic linear motifs, 2014 update. Nucleic Acids Res 42: D259-D266. https://doi.org/10.1093/nar/gkt1047
|
| [25] |
Teilum K, Olsen JG, Kragelund BB (2021) On the specificity of protein–protein interactions in the context of disorder. Biochem J 478: 2035-2050. https://doi.org/10.1042/BCJ20200828
|
| [26] |
Davey NE, Simonetti L, Ivarsson Y (2023) The next wave of interactomics: Mapping the SLiM-based interactions of the intrinsically disordered proteome. Curr Opin Struct Biol 80: 102593. https://doi.org/10.1016/j.sbi.2023.102593
|
| [27] |
Kliche J, Ivarsson Y (2022) Orchestrating serine/threonine phosphorylation and elucidating downstream effects by short linear motifs. Biochem J 479: 1-22. https://doi.org/10.1042/BCJ20200714
|
| [28] |
Correa Marrero M, Mello VH, Sartori P, et al. (2025) Global comparative structural analysis of responses to protein phosphorylation. Nat Commun 16: 9407. https://doi.org/10.1038/s41467-025-64116-4
|
| [29] |
Patel KN, Chavda D, Manna M (2024) Molecular docking of intrinsically disordered proteins: Challenges and strategies. Protein-Protein Docking . Springer US 165-201. https://doi.org/10.1007/978-1-0716-3985-6_11
|
| [30] |
Madeira F, Tinti M, Murugesan G, et al. (2015) 14-3-3-pred: Improved methods to predict 14-3-3-binding phosphopeptides. Bioinformatics 31: 2276-2283. https://doi.org/10.1093/bioinformatics/btv133
|
| [31] |
Chen R, Li X, Yang Y, et al. (2022) Prediction of protein-protein interaction sites in intrinsically disordered proteins. Front Mol Biosci 9: 985022. https://doi.org/10.3389/fmolb.2022.985022
|
| [32] | Vaswani A, Shazeer N, Parmar N, et al. (2017) Attention is all you need. Advances in Neural Information Processing Systems . Curran Associates, Inc. https://doi.org/10.48550/arXiv.1706.03762 |
| [33] |
Baek M, DiMaio F, Anishchenko I, et al. (2021) Accurate prediction of protein structures and interactions using a three-track neural network. Science 373: 871-876. https://doi.org/10.1126/science.abj8754
|
| [34] | Evans R, O'Neill M, Pritzel A, et al. (2021) Protein complex prediction with AlphaFold-multimer. BioRxiv . https://doi.org/10.1101/2021.10.04.463034 |
| [35] |
Bryant P, Pozzati G, Elofsson A (2022) Improved prediction of protein-protein interactions using AlphaFold2. Nat Commun 13: 1175. https://doi.org/10.1038/s41467-022-28865-w
|
| [36] |
Hýsková A, Maršálková E, Šimeček P (2026) Balancing speed and precision in protein folding: A comparison of AlphaFold2, ESMFold, and OmegaFold. Front Genet 16: 1715037. https://doi.org/10.3389/fgene.2025.1715037
|
| [37] |
Obenauer JC, Cantley LC, Yaffe MB (2003) Scansite 2.0: Proteome-wide prediction of cell signaling interactions using short sequence motifs. Nucleic Acids Res 31: 3635-3641. https://doi.org/10.1093/nar/gkg584
|
| [38] |
Dinkel H, Michael S, Weatheritt RJ, et al. (2012) ELM–the database of eukaryotic linear motifs. Nucleic Acids Res 40: D242-D251. https://doi.org/10.1093/nar/gkr1064
|
| [39] |
Kumar M, Michael S, Alvarado-Valverde J, et al. (2024) ELM–the eukaryotic linear motif resource–2024 update. Nucleic Acids Res 52: D442-D455. https://doi.org/10.1093/nar/gkad1058
|
| [40] |
Mooney C, Pollastri G, Shields DC, et al. (2012) Prediction of short linear protein binding regions. J Mol Biol 415: 193-204. https://doi.org/10.1016/j.jmb.2011.10.025
|
| [41] |
Egbert CM, Warr LR, Pennington KL, et al. (2023) The integration of proteome-wide PTM data with protein structural and sequence features identifies phosphorylations that mediate 14-3-3 interactions. J Mol Biol 435: 167890. https://doi.org/10.1016/j.jmb.2022.167890
|
| [42] |
Weesep L van, Özçelik R, Pennings M, et al. (2025) Identifying 14-3-3 interactome binding sites with deep learning. Digit Discov 4: 2602-2614. https://doi.org/10.1039/D5DD00132C
|
| [43] |
Mäeots M-E, Enchev RI (2022) Structural dynamics: Review of time-resolved cryo-EM. Acta Crystallogr D Struct Biol 78: 927-935. https://doi.org/10.1107/S2059798322006155
|
| [44] |
Klebl DP, Aspinall L, Muench SP (2023) Time resolved applications for cryo-EM; approaches, challenges and future directions. Curr Opin Struct Biol 83: 102696. https://doi.org/10.1016/j.sbi.2023.102696
|
| [45] |
Milles S, Salvi N, Blackledge M, et al. (2018) Characterization of intrinsically disordered proteins and their dynamic complexes: From in vitro to cell-like environments. Prog Nucl Magn Reson Spectrosc 109: 79-100. https://doi.org/10.1016/j.pnmrs.2018.07.001
|
| [46] |
Schanda P, Haran G (2024) NMR and single-molecule FRET insights into fast protein motions and their relation to function. Annu Rev Biophys 53: 247-273. https://doi.org/10.1146/annurev-biophys-070323-022428
|
| [47] |
Sebák F, Ecsédi P, Bermel W, et al. (2022) Selective 1Hα NMR methods reveal functionally relevant proline cis/trans isomers in intrinsically disordered proteins: Characterization of minor forms, effects of phosphorylation, and occurrence in proteome. Angew Chem Int Ed 61: e202108361. https://doi.org/10.1002/anie.202108361
|
| [48] |
Sebák F, Szolomájer J, Papp N, et al. (2023) Proline cis/trans isomerization in intrinsically disordered proteins and peptides. Front Biosci (Landmark Ed) 28: 127. https://doi.org/10.31083/j.fbl2806127
|
| [49] |
Mazal H, Haran G (2019) Single-molecule FRET methods to study the dynamics of proteins at work. Curr Opin Biomed Eng 12: 8-17. https://doi.org/10.1016/j.cobme.2019.08.007
|
| [50] |
Hodge EA, Benhaim MA, Lee KK (2020) Bridging protein structure, dynamics, and function using hydrogen/deuterium-exchange mass spectrometry. Protein Sci 29: 843-855. https://doi.org/10.1002/pro.3790
|
| [51] |
Narang D, Lento C, Wilson DJ (2020) HDX-MS: An analytical tool to capture protein motion in action. Biomedicines 8: 224. https://doi.org/10.3390/biomedicines8070224
|
| [52] |
MacArthur MW, Thornton JM (1991) Influence of proline residues on protein conformation. J Mol Biol 218: 397-412. https://doi.org/10.1016/0022-2836(91)90721-H
|
| [53] |
Grathwohl C, Wüthrich K (1981) NMR studies of the rates of proline cis–trans isomerization in oligopeptides. Biopolymers 20: 2623-2633. https://doi.org/10.1002/bip.1981.360201209
|
| [54] |
Wedemeyer WJ, Welker E, Scheraga HA (2002) Proline cis-trans isomerization and protein folding. Biochemistry 41: 14637-14644. https://doi.org/10.1021/bi020461y
|
| [55] |
Reimer U, Fischer G (2002) Local structural changes caused by peptidyl–prolyl cis/trans isomerization in the native state of proteins. Biophys Chem 96: 203-212. https://doi.org/10.1016/s0301-4622(02)00013-3
|
| [56] |
Theisen FF, Prestel A, Jacobsen NL, et al. (2025) Proline cis/trans conformational selection controls 14-3-3 binding. J Am Chem Soc 147: 5714-5724. https://doi.org/10.1021/jacs.4c13462
|
| [57] |
Schmid FX (1993) Prolyl isomerization in protein folding and protein engineering. Annu Rev Biophys Biomol Struct 22: 123-143. https://doi.org/10.1146/annurev.bb.22.060193.001011
|
| [58] |
Yaffe MB, Rittinger K, Volinia S, et al. (1997) The structural basis for 14-3-3-phosphopeptide binding specificity. FEBS Lett 413: 119-124. https://doi.org/10.1016/S0014-5793(97)00826-7
|
| [59] |
Rittinger K, Budman J, Xu J, et al. (1999) Structural analysis of 14-3-3 phosphopeptide complexes identifies a dual role for the amphipathic groove. Mol Cell 4: 153-166. https://doi.org/10.1016/S1097-2765(00)80202-0
|
| [60] |
Schubert M, Labudde D, Oschkinat H, et al. (2002) A software tool for the prediction of xaa-pro peptide bond conformations in proteins based on 13C chemical shift statistics. J Biomol NMR 24: 149-154. https://doi.org/10.1023/a:1020997118364
|
| [61] | Lu KP, Zhou XZ (2007) The prolyl isomerase PIN1: A pivotal new twist in phosphorylation signalling and disease. Nat Rev Mol Cell Biol 8: 904-916. https://doi.org/10.1038/nrm2261 |
Figures(1) / Tables(2)
Sergio Müller, Ignacio Moreno Guerci, Diego Martin Bustos. AI-driven prediction of 14-3-3–intrinsically disordered protein interactions: From SLiM discovery to structural modeling of dynamic binding interfaces[J]. AIMS Biophysics, 2026, 13(4): 424-442. doi: 10.3934/biophy.2026022
DownLoad: