Summary
Structure prediction and design tools trained on monomers generalize to oligomers. ESMFold can predict the structures of some oligomers despite being trained exclusively on monomers (1), and the original released version of AlphaFold2 was trained for monomers but allowed oligomers to be predicted with enough spacing between the sequences for the two chains (ColabFold). On the design side, (2) showed that it is possible to successfully design residues at PPI interfaces using models trained only on monomers, and (3,4) show experimental evidence of successful binder design using ProteinMPNN, which was also only trained on monomers. However, they do not generalize to distinguishing between binders and nonbinders.
1.
Lin Z, Akin H, Rao R, Hie B, Zhu Z, Lu W, et al. Evolutionary-scale prediction of atomic-level protein structure with a language model. Science. 2023;379(6637):1123–30. Available from: https://doi.org/10.1126/science.ade2574
2.
Mahajan SP, Dávila-Hernández FA, Ruffolo JA, Gray JJ. How well do contextual protein encodings learn structure, function, and evolutionary context? Cell Systems. 2025;16(3):101201. Available from: https://doi.org/10.1016/j.cels.2025.101201
3.
Dauparas J, Anishchenko I, Bennett N, Bai H, Ragotte RJ, Milles LF, et al. Robust deep learning–based protein sequence design using ProteinMPNN. Science. 2022;378(6615):49–56. Available from: https://doi.org/10.1126/science.add2187
4.
Vázquez Torres S, Leung PJY, Venkatesh P, Lutz ID, Hink F, Huynh H-H, et al. De novo design of high-affinity binders of bioactive helical peptides. Nature. 2023;626(7998):435–42. Available from: https://doi.org/10.1038/s41586-023-06953-1