For much of history, drug discovery was quite random. However, with the industrial revolution, a systematic approach emerged involving:
Target Identification - What is causing the disease?
Lead Identification & Optimization - What drug molecule can be used to modify the target?
Preclinical Research - Is the drug molecule safe and effective in living organisms like mice?
Clinical Research - Is the drug molecule safe and effective in humans?
Despite the systematic approach, many drugs still fail in clinical trials.
Money-makers don't like this. So, here's where AI comes into play.
Unlike humans, AI is plenty more capable and efficient at sifting through large amounts of data to identify potential targets or leads that we may have missed.
In the past decade, several different AI drug discovery companies were established. So, have they completely revolutionized the drug discovery race?
No, not yet. Clinical research takes a lot of time, and while many different companies do have various drugs in the pipeline or in Phase 1/2 clinical trials, we'll still have to wait awhile before we see the impact of AI on the successful journey of drugs to market.
And as many times as it's been told before, I will reiterate: AI is here to stay. Big pharma has been partnering up or buying out different AI small drug discovery startups, solidifying AI's presence in the drug discovery process.
However, here is my#2cents. No matter what magical tools we use to find 'cures', biology is random and history is random. In other words, AI is founded on historical data archives that were built haphazardly but beautifully by clueless humans, and AI's success in drug discovery depends on whether our biology chooses to cooperate.
So, a sprinkle of spontaneity will forever permeate drug discovery.
2024-04-22
Additional Information:
It’s been a decade of AI in the drug discovery race. What’s next? by Andrii Buvailo
AI in small-molecule drug discovery: a coming wave? by Boston Consulting Group
Comments