Dirichlet latent modelling enables effective learning and sampling of the functional protein design space.
Journal
Nature communications
ISSN: 2041-1723
Titre abrégé: Nat Commun
Pays: England
ID NLM: 101528555
Informations de publication
Date de publication:
29 Oct 2024
29 Oct 2024
Historique:
received:
13
06
2023
accepted:
14
10
2024
medline:
29
10
2024
pubmed:
29
10
2024
entrez:
29
10
2024
Statut:
epublish
Résumé
Engineering proteins with desired functions and biochemical properties is pivotal for biotechnology and drug discovery. While computational methods based on evolutionary information are reducing the experimental burden by designing targeted libraries of functional variants, they still have a low success rate when the desired protein has few or very remote homologous sequences. Here we propose an autoregressive model, called Temporal Dirichlet Variational Autoencoder (TDVAE), which exploits the mathematical properties of the Dirichlet distribution and temporal convolution to efficiently learn high-order information from a functionally related, possibly remotely similar, set of sequences. TDVAE is highly accurate in predicting the effects of amino acid mutations, while being significantly 90% smaller than the other state-of-the-art models. We then use TDVAE to design variants of the human alpha galactosidase enzymes as potential treatment for Fabry disease. Our model builds a library of diverse variants which retain sequence, biochemical and structural properties of the wildtype protein, suggesting they could be suitable for enzyme replacement therapy. Taken together, our results show the importance of accurate sequence modelling and the potential of autoregressive models as protein engineering and analysis tools.
Identifiants
pubmed: 39468034
doi: 10.1038/s41467-024-53622-6
pii: 10.1038/s41467-024-53622-6
doi:
Substances chimiques
alpha-Galactosidase
EC 3.2.1.22
Proteins
0
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
9309Subventions
Organisme : RCUK | Engineering and Physical Sciences Research Council (EPSRC)
ID : EP/V033794/1
Organisme : RCUK | Engineering and Physical Sciences Research Council (EPSRC)
ID : EP/Y01913X/1
Organisme : RCUK | Engineering and Physical Sciences Research Council (EPSRC)
ID : EP/S02431X/1
Informations de copyright
© 2024. The Author(s).
Références
Huang, P.-S., Boyken, S. E. & Baker, D. The coming of age of de novo protein design. Nature 537, 320–327 (2016).
doi: 10.1038/nature19946
pubmed: 27629638
Romero, P. A. & Arnold, F. H. Exploring protein fitness landscapes by directed evolution. Nat. Rev. Mol. Cell Biol. 10, 866–876 (2009).
doi: 10.1038/nrm2805
pubmed: 19935669
pmcid: 2997618
Kiss, Gert, Çelebi-Ölçüm, N., Moretti, R., Baker, D. and Houk, KN. Computational Enzyme Design. Angew. Chem. Int. Ed. 52, 5700–5725. (2013).
Rives, A. et al. Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences. Proc. Natl. Acad. Sci. USA 118, e2016239118 (2021).
doi: 10.1073/pnas.2016239118
pubmed: 33876751
pmcid: 8053943
Kingma, D. P. & Welling, M. Auto-encoding variational bayes. Preprint at https://doi.org/10.48550/arXiv.1312.6114 (2013).
Wu, Z., Johnston, K. E., Arnold, F. H. & Yang, K. K. Protein sequence design with deep generative models. Curr. Opin. Chem. Biol. 65, 18–27 (2021).
doi: 10.1016/j.cbpa.2021.04.004
pubmed: 34051682
Hopf, T. A. et al. Mutation effects predicted from sequence co-variation. Nat. Biotechnol. 35, 128–135 (2017).
doi: 10.1038/nbt.3769
pubmed: 28092658
pmcid: 5383098
Riesselman, A. J., Ingraham, J. B. & Marks, D. S. Deep generative models of genetic variation capture the effects of mutations. Nat. Methods 15, 816–822 (2018).
doi: 10.1038/s41592-018-0138-4
pubmed: 30250057
pmcid: 6693876
Frazer, J. et al. Disease variant prediction with deep generative models of evolutionary data. Nature 599, 91–95 (2021).
doi: 10.1038/s41586-021-04043-8
pubmed: 34707284
Shin, J.-E. et al. Protein design and variant prediction using autoregressive generative models. Nat. Commun. 12, 2403 (2021).
doi: 10.1038/s41467-021-22732-w
pubmed: 33893299
pmcid: 8065141
Repecka, D. et al. Expanding functional protein sequence spaces using generative adversarial networks. Nat. Mach. Intell. 3, 324–333 (2021).
doi: 10.1038/s42256-021-00310-5
Giessel, A. et al. Therapeutic enzyme engineering using a generative neural network. Sci. Rep. 12, 1536 (2022).
doi: 10.1038/s41598-022-05195-x
pubmed: 35087131
pmcid: 8795449
Sjolander, K. et al. Dirichlet mixtures: a method for improved detection of weak but significant protein sequence homology. Bioinformatics 12, 327–345 (1996).
doi: 10.1093/bioinformatics/12.4.327
Bai, S., Kolter, J. Z. & Koltun, V. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling. Preprint at https://doi.org/10.48550/arXiv.1803.01271 (2018).
Platt, F. M., d’Azzo, A., Davidson, B. L., Neufeld, E. F. & Tifft, C. J. Lysosomal storage diseases. Nat. Rev. Dis. Prim. 4, 27 (2018).
doi: 10.1038/s41572-018-0025-4
pubmed: 30275469
Hsu, C., Nisonoff, H., Fannjiang, C. & Listgarten, J. Learning protein fitness models from evolutionary and assay-labeled data. Nat. Biotechnol. 40, 1114–1122 (2022).
doi: 10.1038/s41587-021-01146-5
pubmed: 35039677
Kornreich, R., Desnick, R. J. & Bishop, D. F. Nucleotide sequence of the human alpha-galactosidase A gene. Nucleic Acids Res. 17, 3301 (1989).
doi: 10.1093/nar/17.8.3301
pubmed: 2542896
pmcid: 317741
Parenti, G., Medina, D. L. & Ballabio, A. The rapidly evolving view of lysosomal storage diseases. EMBO Mol. Med. 13, e12836 (2021).
doi: 10.15252/emmm.202012836
pubmed: 33459519
pmcid: 7863408
Xu, S. et al. Coformulation of a novel human α-galactosidase a with the pharmacological chaperone AT1001 leads to improved substrate reduction in fabry mice. Mol. Ther. 23, 1169–1181 (2015).
doi: 10.1038/mt.2015.87
pubmed: 25915924
pmcid: 4817779
Burda, Y., Grosse, R. & Salakhutdinov, R. Importance weighted autoencoders. Preprint at https://doi.org/10.48550/arXiv.1509.00519 (2015).
Li, Y.-D. et al. The rapid evolution of signal peptides is mainly caused by relaxed selection on non-synonymous and synonymous sites. Gene 436, 8–11 (2009).
doi: 10.1016/j.gene.2009.01.015
pubmed: 19393172
Lukas, J. et al. Functional and clinical consequences of novel α-galactosidase A mutations in Fabry disease. Hum. Mutat. 37, 43–51 (2016).
doi: 10.1002/humu.22910
pubmed: 26415523
Skjærven, L., Yao, X.-Q., Scarabelli, G. & Grant, B. J. Integrating protein structural dynamics and evolutionary analysis with Bio3D. BMC Bioinforma. 15, 1–11 (2014).
doi: 10.1186/s12859-014-0399-6
Hallows, W. C. et al. Optimizing human α-galactosidase for treatment of Fabry disease. Sci. Rep. 13, 4748 (2023).
doi: 10.1038/s41598-023-31777-4
pubmed: 36959353
pmcid: 10036536
Hsu, C. et al. Learning inverse folding from millions of predicted structures. in International Conference on Machine Learning, 8946–8970 (2022).
Anishchenko, I. et al. De novo protein design by deep network hallucination. Nature 600, 547–552 (2021).
doi: 10.1038/s41586-021-04184-w
pubmed: 34853475
pmcid: 9293396
Joo, W., Lee, W., Park, S. & Moon, I.-C. Dirichlet variational autoencoder. Pattern Recognit. 107, 107514 (2020).
doi: 10.1016/j.patcog.2020.107514
Mikolov, T., Kombrink, S., Burget, L., Černocky`, J. & Khudanpur, S. Extensions of recurrent neural network language model. in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 5528–5531 (2011).
Hochreiter, S. & Schmidhuber, J. Long short-term memory. Neural Comput. 9, 1735–1780 (1997).
doi: 10.1162/neco.1997.9.8.1735
pubmed: 9377276
Cho, K. et al. Learning phrase representations using RNN encoder-decoder for statistical machine translation. Preprint at https://doi.org/10.48550/arXiv.1406.1078 (2014).
Pascanu, R., Mikolov, T. & Bengio, Y. On the difficulty of training recurrent neural networks. in International Conference on Machine Learning, 1310–1318 (2013).
Wu, Y. et al. Google’s neural machine translation system: Bridging the gap between human and machine translation. Preprint at https://doi.org/10.48550/arXiv.1609.08144 (2016).
Rezende, D. J., Mohamed, S. & Wierstra, D. Stochastic backpropagation and approximate inference in deep generative models. in International Conference on Machine Learning, 1278–1286 (2014).
Figurnov, M., Mohamed, S. & Mnih, A. Implicit reparameterization gradients. Advances in Neural Information Processing Systems 31, (2018).
Jankowiak, M. & Obermeyer, F. Pathwise derivatives beyond the reparameterization trick. in International Conference on Machine Learning, 2235–2244 (2018).
Remmert, M., Biegert, A., Hauser, A. & Soding, J. HHblits: lightning-fast iterative protein sequence searching by HMM-HMM alignment. Nat. Methods 9, 173–175 (2012).
doi: 10.1038/nmeth.1818
Cock, P. J. et al. Biopython: freely available Python tools for computational molecular biology and bioinformatics. Bioinformatics 25, 1422 (2009).
doi: 10.1093/bioinformatics/btp163
pubmed: 19304878
pmcid: 2682512
Lin, Z. et al. Evolutionary-scale prediction of atomic-level protein structure with a language model. Science 379, 1123–1130 (2023).
doi: 10.1126/science.ade2574
pubmed: 36927031
Eastman, P. et al. OpenMM 7: Rapid development of high performance algorithms for molecular dynamics. PLoS Comput. Biol. 13, e1005659 (2017).
doi: 10.1371/journal.pcbi.1005659
pubmed: 28746339
pmcid: 5549999
Grant, B. J., Skjærven, L. & Yao, X.-Q. The Bio3D packages for structural bioinformatics. Protein Sci. 30, 20–30 (2021).
doi: 10.1002/pro.3923
pubmed: 32734663