|
ChatGPT For Drug Development?
(Thursday, March 9, 2023) Artificial Intelligence systems such as ChatGPT can be used for creating summaries, discussions, and generating ideas. Can that be expanded to designing new drugs and finding new uses or previously approved drugs? ChatGPT uses publicly available information to summarize and pseudo-analyze specific questions posed to it providing seemingly intellectual assessments. The software can theoretically be trained to go further and help create new drugs. One can envision an AI system that combines the information available externally with that generated internally to draw conclusions, the way ChatGPT does in essay writing or discussion papers. An interesting article in MIT Tech Review last month presented such a scenario. AI was used to design drugs that are now in clinical trials. And this is not an isolated incidence. There are hundreds of companies involved in creating AI-based systems for drug design. In fact, the idea is not new. In-silico models for designing and testing new drugs have been used for almost two decades. “With machine learning, vast amounts of data, including drug and molecular data, can be harnessed to build complex models automatically. This makes it far easier—and faster—to predict how drugs might behave in the body, allowing many early experiments to be carried out in silico. Machine-learning models can also sift through vast, untapped pools of potential drug molecules in a way that was not previously possible.” ChatGPT has generated intense interest by providing access to this technology to everyone. The applications of ChatGPT such as text generation, answering everyday questions, and performing daily mundane tasks are just the tip of the iceberg. AI systems can be used effectively for selection of targets for the drugs such as a biological pathway, an antigen, a symptom, etc. It can also be used to identify the overall patient characteristics for the drug, very much like defining the inclusion and exclusion criteria to select patients potentially most likely to be helped by a new drug. Then it can help design small or large molecules for interacting with the target. There are limitations to using AI systems. ChatGPT for one is not (yet!) able to address legal and ethical issues, and lacks the ability for critical thinking, two crucial aspects of drug development. And you still need to do the non-clinical and clinical testing of the drugs designed using AI systems. The big difference between in silico testing from years ago and the technology available today is that the computational resources available now for big data analysis are far superior to that available even 10 years ago. Only time will tell if this is just like the past fad for in silico drug design which faded away from the hype almost two decades ago, or will it become the new norm for drug development. 
AUTHOR
Dr. Mukesh Kumar Founder & CEO, FDAMap Email: [email protected] Linkedin: Mukesh Kumar, PhD, RAC Instagram: mukeshkumarrac Twitter: @FDA_MAP Youtube: MukeshKumarFDAMap |
|
|
|
