Deep learning to design nuclear-targeting abiotic miniproteins.

Autor: Schissel CK; Department of Chemistry, Massachusetts Institute of Technology, Cambridge, MA, USA., Mohapatra S; Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA., Wolfe JM; Department of Chemistry, Massachusetts Institute of Technology, Cambridge, MA, USA.; Ra Pharmaceuticals, Cambridge, MA, USA., Fadzen CM; Department of Chemistry, Massachusetts Institute of Technology, Cambridge, MA, USA.; Harvard Medical School, Boston, MA, USA., Bellovoda K; Sarepta Therapeutics, Cambridge, MA, USA., Wu CL; Sarepta Therapeutics, Cambridge, MA, USA., Wood JA; Sarepta Therapeutics, Cambridge, MA, USA., Malmberg AB; Sarepta Therapeutics, Cambridge, MA, USA., Loas A; Department of Chemistry, Massachusetts Institute of Technology, Cambridge, MA, USA., Gómez-Bombarelli R; Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA, USA. rafagb@mit.edu., Pentelute BL; Department of Chemistry, Massachusetts Institute of Technology, Cambridge, MA, USA. blp@mit.edu.; The Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology, Cambridge, MA, USA. blp@mit.edu.; Center for Environmental Health Sciences, Massachusetts Institute of Technology, Cambridge, MA, USA. blp@mit.edu.; Broad Institute of MIT and Harvard, Cambridge, MA, USA. blp@mit.edu.
Jazyk: angličtina
Zdroj: Nature chemistry [Nat Chem] 2021 Oct; Vol. 13 (10), pp. 992-1000. Date of Electronic Publication: 2021 Aug 09.
DOI: 10.1038/s41557-021-00766-3
Abstrakt: There are more amino acid permutations within a 40-residue sequence than atoms on Earth. This vast chemical search space hinders the use of human learning to design functional polymers. Here we show how machine learning enables the de novo design of abiotic nuclear-targeting miniproteins to traffic antisense oligomers to the nucleus of cells. We combined high-throughput experimentation with a directed evolution-inspired deep-learning approach in which the molecular structures of natural and unnatural residues are represented as topological fingerprints. The model is able to predict activities beyond the training dataset, and simultaneously deciphers and visualizes sequence-activity predictions. The predicted miniproteins, termed 'Mach', reach an average mass of 10 kDa, are more effective than any previously known variant in cells and can also deliver proteins into the cytosol. The Mach miniproteins are non-toxic and efficiently deliver antisense cargo in mice. These results demonstrate that deep learning can decipher design principles to generate highly active biomolecules that are unlikely to be discovered by empirical approaches.
(© 2021. The Author(s), under exclusive licence to Springer Nature Limited.)
Databáze: MEDLINE
Externí odkaz: