Young woman researcher
(Photo : Bigstock)

Over the past several years, the topic of artificial intelligence-enabled automation has received plenty of attention. Most of that attention, quite predictably, has centered on what impacts the technology would have on today's workforce. Experts have written extensively on the subject, with most predicting mass displacement of workers and an attendant need to re-engineer the global economy to account for it.

Lost in the doom-and-gloom headlines, however, is the fact that automation and AI also have the power to create vast benefits for humanity. Nowhere is that truer than in the field of pharmacology. It's a discipline that's gone mostly unchanged for decades, with drug discovery operations boasting high costs and even higher failure rates. It is into that landscape that the latest technological innovations are now advancing, bringing new hope of a more efficient, more successful future.

At this stage, the only real question on the subject left unanswered is how close we are to a time when drug discovery processes will become fully automated and free from human intervention. 

Automation in Drug Discovery Isn't New

In looking at how automation is evolving within the drug discovery process, it's important to note that it's nothing new for most research facilities. Many (if not most), have long made use of various high throughput screening systems to accelerate their lead generation processes, eliminating the repetitive manual labor required to narrow down compound candidates from vast chemical libraries.

Such systems, while faster and more efficient than relying on human researchers conducting individual experiments, still represent something of a hit-or-miss approach. They only serve to identify leads that might prove promising in treating a specific target, but don't give any indication as to toxicity or potential side effects the leads might have. That leaves a great deal of manual work left in the process of developing a new drug.

AI Filling in the Gaps

In recent years, there's been a great deal of development in the field of AI aimed at accelerating research into new drugs and related therapies. Much of that development has been in systems designed to further reduce the time it takes to get from target identification to clinical trials. In some cases, AI platforms are already proving adept at choosing leads with high odds of success, bypassing the traditional screening process altogether.

Other AI systems work to deduce the potential for toxicity of a given lead, further narrowing results and eliminating leads that would eventually fail later in the process. The biggest issue these kinds of solutions have encountered so far is the variety and accuracy of the data sets they rely on to work. Some approaches are already using advanced computational molecule modeling to create training data from scratch to make up for lacking information. It's too soon, though, to know how reliable that process will be - as the platforms have no track record to evaluate.

When End-to-End Automation Will Come

Based on the current state of automation and AI technology as it's been applied to drug development, it's easy to estimate that fully-automated drug discovery processes could go mainstream within the next year or two. They aren't likely to displace traditional drug discovery methods anytime soon, though.

For that to happen, a successful drug would have to make it to market in order to create the industry momentum towards wholesale process change. There's already one such example moving into clinical trials, however. It's an AI-developed drug aimed at treating obsessive compulsive disorder. The AI behind it used patient genetic data as a roadmap to identifying the molecule that's now the basis for the trial. That means we could be around seven years away from seeing a successful, autonomously-developed drug reach the approval stage, laying the foundation for a new generation of medications.

The Bottom Line

It's clear that the technology needed to support an end-to-end automated drug discovery solution is on the verge of being ready for prime time. Even major governments are getting in on the act, with the UK about halfway through construction of a research institute geared toward developing "fully-automated hands-free molecular discovery" for next-generation drug development. With major backing from venture capitalists, academia, and the pharmaceutical industry, it's a safe bet that an end-to-end solution is right around the corner, and will change everything about medicine as we know it today - with benefits we can't yet imagine.