Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules
[1];
[2];
[2];
[6]
- Kyulux North America Inc., 10 Post Office Square, Suite 800, Boston, Massachusetts 02109, United States
- Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts 02138, United States
- Department of Computer Science, University of Toronto, 6 King’s College Road, Toronto, Ontario M5S 3H5, Canada
- Department of Engineering, University of Cambridge, Trumpington Street, Cambridge CB2 1PZ, U.K.
- Google Brain, Mountain View, California, United States, Princeton University, Princeton, New Jersey, United States
- Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts 02138, United States, Biologically-Inspired Solar Energy Program, Canadian Institute for Advanced Research (CIFAR), Toronto, Ontario M5S 1M1, Canada
We report a method to convert discrete representations of molecules to and from a multidimensional continuous representation. This model allows us to generate new molecules for efficient exploration and optimization through open-ended spaces of chemical compounds. A deep neural network was trained on hundreds of thousands of existing chemical structures to construct three coupled functions: an encoder, a decoder, and a predictor. The encoder converts the discrete representation of a molecule into a real-valued continuous vector, and the decoder converts these continuous vectors back to discrete molecular representations. The predictor estimates chemical properties from the latent continuous vector representation of the molecule. Continuous representations of molecules allow us to automatically generate novel chemical structures by performing simple operations in the latent space, such as decoding random vectors, perturbing known chemical structures, or interpolating between molecules. Continuous representations also allow the use of powerful gradient-based optimization to efficiently guide the search for optimized functional compounds. We demonstrate our method in the domain of drug-like molecules and also in a set of molecules with fewer that nine heavy atoms.
- Research Organization:
- Harvard Univ., Cambridge, MA (United States); Univ. of Toronto, ON (Canada)
- Sponsoring Organization:
- USDOE Office of Science (SC), Basic Energy Sciences (BES)
- Grant/Contract Number:
- SC0015959
- OSTI ID:
- 1416858
- Alternate ID(s):
- OSTI ID: 1498675
- Journal Information:
- ACS Central Science, Journal Name: ACS Central Science Vol. 4 Journal Issue: 2; ISSN 2374-7943
- Publisher:
- American Chemical SocietyCopyright Statement
- Country of Publication:
- United States
- Language:
- English
Web of Science
Similar Records
Bayesian optimization in continuous spaces via virtual process embeddings
Deep learning to generate in silico chemical property libraries and candidate molecules for small molecule identification in complex samples