Credits: Social Media
Scientists are using AI to find hundreds of thousands of new psychedelic drugs
Scientists are turning to AI to find hundreds of thousands of new potential psychedelic drugs to use as antidepressants.
Over the past 10 years, researchers from around the world have been busy promoting the use of psychedelic drugs to treat things like depression, anxiety, and PTSD with varying degrees of success. As the race to find the next generation of medicinal psychedelics heats up, scientists are now turning to AI to help find hundreds of thousands of possible new drugs.
Scientists turn to AI to help find new psychedelic drugs
Scientists have used an AI tool called AlphaFold to identify what they claim to be hundreds of thousands of potential new psychedelic molecules that could be used as antidepressants.
The idea is that if AI has access to an enormous database of known molecules, it could be used by the pharmaceutical industry to find new and effective ways to treat diseases and other ailments.
This is according to scientists from the University of North Carolina at Chapel Hill, who recently teamed up with British Artificial Intelligence company DeepMind, to publish a preprint study in late 2023.
Incredibly, but certainly not without skepticism (see next section), the team found that AI-predicted structures had a 60% hit rate when it came to identifying drugs for a sought-after class of target molecules.
What makes this research noteworthy for you and me? This study demonstrates how AI can help identify new drugs at the touch of a button, instead of the months or even years that it would take using human scientists.
AlphaFold is an absolute revolution. If we have a good structure, we should be able to use it for drug design, celebrated Jens Carlsson, a chemist from Swedens University of Uppsala.
Theres a lot of hype, but some scientists D.A.R.E to have more skepticism
While there is understandably a lot of excitement surrounding the use of AI in medicine, scientists are acutely aware that the general public should have more skepticism before jumping on the next psychedelic bandwagon.
As University of California pharmaceutical chemist Brian Shoichet, one of the authors of the paper, explained to Nature:
There is a lot of hype. Whenever anybody says, such and such is going to revolutionize drug discovery, it warrants some skepticism.
In fact, Shoichet shared how there are more than 10 peer-reviewed studies out there that have found AlphaFolds results to be less useful than traditional scientific methods. Yet curiosity pushed him to see if he could replicate such results, with Shoichets team synthesizing hundreds of compounds from both traditional human and AI-predicted models.
- BEST AND BRIGHTEST: US astronaut reveals what he missed most about life on Earth
Shoichet said that, while no two molecules that were the same, [and] they didnt even resemble each other , the hit rate between the two groups were almost identical, which he said was a genuinely new result.
Ultimately, Shoichet came to the conclusion that AI predictions will not be universally useful, but also that compared to actually going out and getting a new structure, you could advance the project by a couple of years and thats huge.
So, while AI is unlikely to replace traditional experimentation in drug discovery anytime soon, the researchers did note that the value of using AI tools to find new drugs like psychedelics should be recognized.
We asked ChatGPT about using AI software to find new psychedelics
OpenAIs incredibly popular ChatGPT software is certainly not on the same level as the AlphaFold system, nor can it be considered a legitimate source of authority when it comes to this type of research.
That being said, what did the program say when we asked about the use of AI to find new psychedelic drugs for such research purposes?
At first, ChatGPT gave a generic answer about the interesting and potentially promising idea but gave a surprisingly honest assessment of the risks and benefits when prompted for a more detailed answer.
The software noted that, aside from targeted design, using AI here could significantly speed up the drug discovery process.
ChatGPT also identified that this could reduce both costs and time while making the field more accessible to scientists from less privileged countries around the world.
However, the platform also identified that there are safety concerns, ethical considerations, and regulatory hurdles that need to be addressed before such research can be brought into the real world.
The potential benefits of advancing our understanding of psychedelics should be balanced with a commitment to ensuring the responsible and ethical development of these substances.
The original research paper was published via bioRxiv as part of the pre-print server; the paper has not yet been peer-reviewed.
