Faculty Opinions recommendation of Deep learning enables rapid identification of potent DDR1 kinase inhibitors.
Problem
Discovering potent inhibitors of DDR1 kinase, a target implicated in fibrosis and other diseases, is slow using traditional methods.
Why it matters
Faster drug discovery for fibrosis and related diseases could reduce suffering and treatment costs.
What the AI did
A deep generative model (GENTRL) designed novel molecules optimizing synthetic feasibility, novelty, and biological activity, discovering potent DDR1 inhibitors in 21 days.
Results
Four compounds active in biochemical assays; two validated in cell-based assays; one lead candidate showed favorable pharmacokinetics in mice.
Impact
AI-generated compounds demonstrated biological activity and favorable pharmacokinetics in preclinical testing.
Record
| Venue | Faculty Opinions – Post-Publication Peer Review of the Biomedical Literature |
|---|---|
| Published | 2019 |
| Work type | peer-review |
| Authors | Peter R Bernstein |
| Citations | 0 (OpenAlex, at import time) |
| Model or method | GENTRL (generative tensorial reinforcement learning) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3410/f.736535355.793566016 |
| Open access | closed |
| Keywords | GENTRL, drug discovery, DDR1 kinase, generative model, de novo design |
Importance rating
real but narrower advance
Accelerates drug discovery for a specific kinase target implicated in fibrosis; abstract gives no scale of disease burden, so impact is a real but narrower advance in preclinical drug discovery.