Surfing the wave between the wet lab and AI-driven technologies

I was young, maybe 8 or 9. I don’t remember who, maybe my mother or grandmother, but someone gave me a book about evolution. It was called The Story of Life opens in new tab/window, and the illustrations were fascinating. Soon after, as kids often do, I got hooked on dinosaurs. Since then, I’ve always loved biology. My parents tried to suggest, ‘How about becoming a doctor?’ But no, I was stuck to biology.

John Vlahos opens in new tab/window is reminiscing about his formative days in Greece. Today, he lives in New York, where he’s been working as a Senior Associate Scientist at the biotech company Regeneron opens in new tab/window for almost two years. During this time, he also got stuck on data science, for which he is pursuing a master’s degree.

With his unique blend of wet lab expertise and data science skills, John embodies the changing landscape of pharmaceutical research in the AI era — bridging traditional biological research with sophisticated data analytics.

John is one of the guests featured in Elsevier’s webinar AI in action: R&D perspectives on tools, myths and real-world impact opens in new tab/window. He’s joined by representatives from Cargill and ABB.

A passion for science

Before joining Regeneron, John worked at NYU Langone Health opens in new tab/window, where he researched peripheral artery disease and aneurysms. Now, he focuses on cardiovascular disease research, specifically blood coagulation disorders and hemophilia: “It’s exciting,” he says. “Coagulation is quite intricate as a pathway molecular biological system.” His day-to-day work involves testing potential therapeutic compounds:

My job primarily is to find new targets for therapies. I’m on the bench every day. I take whatever the chemists make and do many in vitro and animal tests. I create reports and see if the drug works, how it works, and if it’s safe.

He says he values the company’s vibrant culture and scientific focus: “The people here really care about the science. They’re very passionate about the work they do. With its atmosphere of curiosity and exploration, it feels more like a university while still being a company out to compete. My colleagues and I are very focused on fulfilling our mission of delivering much-needed therapeutics to the people who really need them.”

Bridging science with technology

John recognizes AI’s potential to transform pharmaceutical research while also acknowledging the challenges of implementation. “For some biologists, it’s difficult to see the potential value of technologies like artificial intelligence,” John observes. “They’re more traditional scientists — they prefer Excel spreadsheets and reports, which are robust but slow.”

John says he has become an unofficial translator between scientific and technological domains, helping colleagues understand how AI tools can enhance rather than replace traditional methods.

He sees a connection between this role and his job as a medical translator during the pandemic: “There’s a big Greek community here in New York City, and a lot of the older people don’t speak English very well. And when they had to go to the doctor, I bridged the parlance of the doctors to them. It was a very rewarding experience — getting to help people who reminded me of my grandparents. But I also learned to communicate between different types of audiences. This proved to be an important skill for me.”

AI in daily research

As an early adopter and “heavy user” of Elsevier’s AI research tools — most recently ScienceDirect AI — John has experienced firsthand how AI can transform scientific workflows. What previously took three to four hours of literature searching can now be accomplished in 30 minutes through natural language queries and AI-powered analysis:

The potential for AI in pharma R&D

John sees significant potential for AI in pharmaceutical R&D. While taking a cautious approach to external AI tools — Regeneron has restricted access to popular LLMs due to data privacy and intellectual property concerns — the company is developing internal AI solutions.

“We have a huge genetic database opens in new tab/window. It’s one of the largest in the world, representing so much usable human data for drug discovery,” John says.

Meanwhile, he adds, Elsevier’s AI tools are constantly being improved: “They’re getting better and better. They just keep advancing.”

AI might not solve everything, but ...

However, the technological developments interest John less than the related conversations.

“I appreciate the debate between the enthusiasts and, let’s say, the pessimists,” he says. “Many people say, ‘This is going to be great’ — that artificial general intelligence will solve all the problems. And then there are people on the other side who say, ‘This is just a chat tool’ — that it’s just a hype thing. I think the truth is somewhere in the middle.

“Of course, there’s a lot of hype around it. All technologies come with hype. But I see the potential. It might not be as great as people say, but it will have a huge impact, even if it ends up being more in the shadows.”

John sees the real value in practical, everyday applications: “Basically, it’s a knowledge condenser that can expedite your own work.”

Yet, he remains realistic about AI’s limitations. “People say it’s not thinking or understanding — that it’s just regurgitating. And this can be true. You need to stay very skeptical about what it does. You can’t treat it like a religious text.” Like evolution, he believes it will take some time before everyone accepts it.

Leading by example

John has clear ideas about how to introduce AI to more traditionally minded colleagues. “It needs to be gradual, and you must demonstrate clear benefits,” he says.

“You can recruit certain individuals to start using it and evangelizing about it — I use this and that and it helps me in these ways. And when people see someone doing their work twice as fast as they are, they will naturally become curious. They will wonder, ‘How is this possible?’

“To anyone, I would say to try it for 10 minutes every day. If it wins you over, what’s the bad thing? You just have another tool in your belt.”

Future vision: accelerating improved outcomes

John anticipates that AI will profoundly influence drug discovery and clinical trials, aiding researchers in comprehending the reasons for drug success or failure. He also imagines a future featuring specialized AI tools tailored for specific research tasks instead of one-size-fits-all solutions.

He also acknowledges that automation and AI will likely transform traditional wet lab work. “I think bench work is going to be more of a hobby, or maybe something done in universities,” he reflects. “The future will be less about work on the bench and more about thinking, creative analysis and ideas.”

Like evolution, transforming workflows often takes time. However, with researchers like John, it can only accelerate — especially since it’s all about accelerating the development of life-changing therapies.