Journal
JOURNAL OF CHEMINFORMATICS
Volume 13, Issue 1, Pages -Publisher
BMC
DOI: 10.1186/s13321-021-00494-3
Keywords
Generative Adversarial Network; Drug discovery; Search
Categories
Funding
- U.S. Department of Energy, Office of Science, through the Office of Advanced Scientific Computing Research (ASCR) [DE-AC05-00OR22725]
- Exascale Computing Project (ECP), of the U.S. Department of Energy Office of Science [17-SC-20-SC]
- Exascale Computing Project (ECP), of the National Nuclear Security Administration [17-SC-20-SC]
- Joint Design of Advanced Computing Solutions for Cancer (JDACS4C) program
- U.S. Department of Energy by Argonne National Laboratory [DE-AC02-06-CH11357]
- Lawrence Livermore National Laboratory [DE-AC52-07NA27344]
- Los Alamos National Laboratory [DE-AC5206NA25396]
- Oak Ridge National Laboratory [DE-AC05-00OR22725]
- Frederick National Laboratory for Cancer Research [HHSN261200800001E]
Ask authors/readers for more resources
This study introduces a new approach to training GANs using concepts from Genetic Algorithms to promote incremental exploration and limit the impacts of mode collapse, outperforming traditional methods by replacing training data samples with valid samples from the generator, leading to a significant increase in potential applications for GANs in drug discovery.
The process of drug discovery involves a search over the space of all possible chemical compounds. Generative Adversarial Networks (GANs) provide a valuable tool towards exploring chemical space and optimizing known compounds for a desired functionality. Standard approaches to training GANs, however, can result in mode collapse, in which the generator primarily produces samples closely related to a small subset of the training data. In contrast, the search for novel compounds necessitates exploration beyond the original data. Here, we present an approach to training GANs that promotes incremental exploration and limits the impacts of mode collapse using concepts from Genetic Algorithms. In our approach, valid samples from the generator are used to replace samples from the training data. We consider both random and guided selection along with recombination during replacement. By tracking the number of novel compounds produced during training, we show that updates to the training data drastically outperform the traditional approach, increasing potential applications for GANs in drug discovery.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available