The following article is an opinion piece written by Andrew Radin. The views and opinions expressed in this article are those of the author and do not necessarily reflect the official position of Technology Networks.
Artificial intelligence (AI) is becoming more widely adopted in the pharmaceutical industry, creating both excitement and questions about its potential and long-term success. In recent years, a plethora of companies – ranging from large pharmaceuticals to startups – have set expectations of AI as a panacea that will revolutionize the industry. While the thought of AI as such is an appealing one, it is not realistic.
Investors have largely gravitated toward this hype throwing an unprecedented amount of funding toward AI-focused startups. However, the immediate expectation of new therapies against intractable diseases have yet to be met. As such, we are seeing a wave of devaluations, decreased investing, and heavy disappointment toward the industry.
There is a reason for this. We simply can’t deploy a machine alone to find new treatments against complex human biology. The science of drug discovery is rigorous and as such our expectations need to be realigned. The practical implementation of AI and how we think about it combined with drug discovery is going to be an evolution, not a revolution. AI’s place in discovery is a complex relationship that needs to be handled with care, it is by no means a panacea.
AI is a branch of computer science designed to mimic how the human brain solves problems and makes decisions. It has been around for nearly 100 years and the use of AI is not what’s new in the story of pharmaceutical innovation. The sophistication of drug discovery has been evolving over decades, and in fact, AI has been used to help support that evolution, although it’s not widely discussed. A classic example is using AI models to help to determine relationships between the structural properties of chemical compounds and biologic activity. They are essential for drug discovery and help scientists better predict how a drug candidate will work in the body. While their predictions are limited by the constraints of the models, they have introduced great efficiency into the drug discovery process, allowing scientists to focus on potential drugs that have an increased chance in fighting a particular disease.
Today, however, we are trying to solve for much more complex diseases and trying to fight them with more accuracy, safety and efficacy than the previous treatments that came before. Fortunately, we are now in an era containing a wealth of data about human biology on top of the ability to analyze large amounts of that data thanks to inexpensive and powerful technology. The potential of AI to address these complex diseases has greatly increased, with the caveat that so has the difficulty of finding treatments and cures.
We can now build an entire virtual world around drug discovery, including in-silico models that simulate human disease using vast amounts of genomic, phenotypic and chemistry data. That data can be accessed freely and analyzed inexpensively. We can use computational methods and algorithms to identify disease features that discovery methods commonly miss due to their dependence on a single predefined hypothesis. We can evaluate potential treatments against multiple targets at the same time. We, as humans, can only do one thing at a time. AI fills that gap for us, but it still needs us to guide it along the way.
We can increase the speed at which drugs are brought into preclinical testing by using AI to truncate the steps needed to begin in vivo testing. We can cross-reference libraries of potential compounds against disease targets at lightning speed. We can now better predict the viability of these compounds against safety and efficacy markers. This level of progress would take years using the traditional methods but by integrating technology, against just one or two compounds, we can do all of this in a matter of weeks.
Steady investments toward AI are paying off. However, unrealistic expectations are unintentionally creating barriers for even wider adoption. Multiple AI companies have identified novel new treatments against new disease targets that hold great potential to treat previously untreatable diseases, including lupus, glioblastoma, aggressive cancers, and fibrotic diseases. The fact that we can use AI to increase the speed of finding these new potential treatments is a huge success – it is creating a pipeline of new drugs that may soon change how we address disease.
AI is already impacting drug discovery in new and previously unimaginable ways. But it’s all a matter of how we judge success. If it’s a machine alone curing a complex disease, then we’ll never achieve success.
The potential of AI multiplies when paired with better education, because with a better understanding of the possibilities and expectations more adoption will happen. The more companies we have improving drug discovery with AI, the more treatments we’ll find over time. However, those candidates still need to endure years of clinical research and prove that they are safe and effective in humans. While we may have changed the timeline aggressively through the proper application of AI, we still have a roadmap to follow that will take years and require rigorous scientific work.
It is going to be an evolving science. The strongest players will continue to build a steady stream of results, even if they come more slowly and with less fanfare than founders, investors and the media had wanted. This steady stream of empirical evidence will drive a new appreciation for AI. One where true value is delivered.
Any scientist working in an R&D lab will tell you that they have harnessed all the technology available to its highest potential in the spirit of treating diseases and improving lives. For us to say AI will revolutionize their work is a disservice to all the innovation that came before us. We need to continue treating it as an evolution that will take place over time and understand how far we’ve already come.
The fact remains that AI has been used for decades, evolving along with the availability of more powerful computational power and data availability. This will continue and we will discover more breakthroughs as a result. Those breakthroughs won’t happen overnight, but they will happen.
About the author:
Andrew A. Radin is the co-founder and chief executive officer of Aria Pharmaceuticals. Andrew created the company’s first drug development algorithms as part of his studies in biomedical informatics at Stanford University in 2014. Since co-founding Aria, Andrew was named an Emerging Pharma Leader by Pharma Executive magazine, was invited to give a TEDMED talk, and was named a Top 100 AI Leader by Deep Knowledge Analytics. In addition to his duties as CEO at Aria, Andrew serves as an advisor to Stanford University’s SPARK drug development and Stanford University’s StartX startup accelerator programs. Prior to co-founding Aria, Andrew served as chief technology officer at several successful internet startups. His prior projects have reached tens of millions of people in telephony systems, advertising networks, video games and geographical mapping systems. Andrew studied Biomedical Informatics at Stanford University’s SCPD graduate program and holds a Master of Science and Bachelor of Science degrees in Computer Science from Rochester Institute of Technology.