Abstract | ||
---|---|---|
In drug discovery, one of the most important tasks is to find novel and biologically active molecules. Given that only a tip of iceberg of drugs was founded in nearly one-century's experimental exploration, it shows great significance to use in silico methods to expand chemical database and profile drug-target linkages. In this study, a web server named ChemGenerator was proposed to generate novel activates for specific targets based on users' input. The ChemGenerator relies on an autoencoder-based algorithm of Recurrent Neural Networks with Long Short-Term Memory by training of 7 million of molecular Simplified Molecular-Input Line-Entry System as the basic model, and further develops target guided generation by transfer learning. As results, ChemGenerator gains lower loss (<0.01) than existing reference model (0.2 similar to 0.4) and shows good performance in the case of Epidermal Growth Factor Receptor. Meanwhile, ChemGenerator is now freely accessible to the public by http://smiles.tcmobile.org. In proportion to endless molecular enumeration and time-consuming expensive experiments, this work demonstrates an efficient alternative way for the first virtual screening in drug discovery. |
Year | DOI | Venue |
---|---|---|
2021 | 10.1093/bib/bbaa407 | BRIEFINGS IN BIOINFORMATICS |
Keywords | DocType | Volume |
deep learning, autoencoder, long short-term memory (LSTM), molecular generation, web server | Journal | 22 |
Issue | ISSN | Citations |
4 | 1467-5463 | 0 |
PageRank | References | Authors |
0.34 | 0 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jing Yang | 1 | 41 | 11.98 |
Ling Hou | 2 | 5 | 1.09 |
Kun-Meng Liu | 3 | 0 | 0.34 |
Wen-Bin He | 4 | 0 | 0.34 |
Yong Cai | 5 | 0 | 0.34 |
Feng-Qing Yang | 6 | 0 | 0.34 |
Yuanjia Hu | 7 | 0 | 1.01 |