A Systematic Review of Deep Learning Methodologies Used in the Drug Discovery Process with Emphasis on In Vivo Validation
Problem
Drug discovery is long and costly, and deep learning results are often not confirmed experimentally before impacting development.
Why it matters
Abstract states drug development is extremely long and costly and computational results often go unvalidated.
What the AI did
A systematic review of deep learning architectures in drug discovery, covering studies where DL-generated or identified molecules/peptides were biologically evaluated in animal models.
Results
Deep learning generated/identified molecules biologically evaluated in animal models; no quantitative metrics in abstract.
Impact
Documents that some DL-designed molecules reached in vivo validation, but no approved drug is claimed.
Record
| Venue | International Journal of Molecular Sciences |
|---|---|
| Published | 2023 |
| Work type | review |
| Authors | Nikoletta-Maria Koutroumpa, Konstantinos D. Papavasileiou, Anastasios G. Papadiamantis, Georgia Melagraki, Antreas Afantitis |
| Citations | 35 (OpenAlex, at import time) |
| References | 92 |
| Model or method | not stated in source |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3390/ijms24076573 |
| Open access | gold |
| Keywords | deep learning, drug discovery, in vivo validation, systematic review, molecular design |
Importance rating
major — millions of lives, an industry or a field
Abstract states drug development is extremely long and costly and computational results often go experimentally unconfirmed; the field affects an entire industry and millions of patients, though no numbers are given.