Can Relay Therapeutics, Aiming to Revolutionize Drug Design, Continue the Path Pioneered by Vertex?

When two proteins are frozen into a lattice-like structure, they look almost identical. For scientists trying to design a drug that locks onto only one of them without affecting the other, this similarity poses a major challenge.

Relay Therapeutics, headquartered in Cambridge, Massachusetts, is a biotechnology company founded just five years ago that has raised nearly $1 billion, with the promise of having the right tools to solve this problem.

With powerful computing capabilities and a range of newer drug discovery technologies, Relay is able to analyze protein motion and conformational changes. The company's CEO, Sanjiv Patel, believes these "molecular movies" help Relay distinguish between similar proteins, enabling the design of better drugs in less time.

This year, Relay unveiled its pipeline, disclosing its first three investigational drugs. Each targets a type of cancer, marking a step in Relay's continuation of a drug design journey that began in 1989 with biotechnology pioneer Vertex Pharmaceuticals.

"We are the result of what Vertex wanted to create 30 years ago," Patel said in an interview.

Vertex and other companies had attempted to design more effective drugs by understanding the three-dimensional structures of disease-causing proteins. They reasoned that understanding the structure of the lock helps design a better key. This "rational" approach was a major advance over earlier, less precise methods.

However, what was missing from the images they constructed was a detailed understanding of how proteins move. Proteins are never static in their natural state; they sway, twist, coil, and fold—more like shape-shifting putty than locks.

"Everyone in structure-guided drug discovery has long recognized that their targets are not static," said Joshua Boger, founder of Vertex and its CEO for two decades. "Whatever method you use to understand biological processes, or to design a drug that treats a protein as a fixed module, is making an extreme assumption."

Relay's claim is that it doesn't have to make that assumption. Its founding team includes longtime Vertex scientist Mark Murcko and billionaire hedge fund manager David Shaw. Relay appears to have already won over Wall Street. The company went public in July, raising $425 million through its initial public offering, making it one of the larger biotech IPOs of the year. Its stock surged 75% on the first day of trading and has remained at that level since.

Although investors may be convinced, proving that protein motion leads to better drugs is much more difficult.

"Protein motion may be important, but there are many other factors that could be equally or more important," said Derek Lowe, a longtime drug discovery researcher, in an interview. "How do you know this is the knife needed to cut through this problem?"

"Like looking at photos of identical twins"

Relay discovered that these two seemingly identical proteins do not move in the same way. A flap-like structure on one of them is looser and extends outward from the amino acid tangle more frequently than on the other.

Both belong to the fibroblast growth factor receptor protein family. Alterations in one of them, called FGFR2, are believed to play a role in many cancers, making it an ideal target for drug developers.

However, blocking only FGFR2 is difficult. Two recently approved anti-FGFR drugs—Incyte's Pemazyre and Johnson & Johnson's Balversa—also inhibit FGFR1 activity, which can lead to elevated blood phosphorus levels. As a result, patients using these drugs require close monitoring and sometimes dose reductions, which Relay believes limits the drugs' effectiveness.

"If you design a drug that inhibits the entire family, it causes a lot of toxicity," said Don Bergstrom, head of R&D at Relay, in an interview. "No one has really been able to use selective drugs because it's like looking at photos of identical twins."

Such photos are the starting point for most drug developers, and Relay is no exception. Typically obtained by bombarding crystallized proteins with X-rays, these images can reveal "pockets" in protein structures that researchers then target with compounds.

At Relay, researchers input these snapshots into computer models that simulate how proteins might move. Their work benefits from supercomputers built by D.E. Shaw Research, a company founded by Shaw after he left his investment firm.

D.E. Shaw's supercomputer, named Anton 2, is crucial to Relay, enabling it to predict the behavior of hundreds of thousands of protein atoms over time spans far beyond what other methods can achieve.

Computer models of protein motion led Relay to realize that FGFR1 moves differently from FGFR2, and the biotech company hopes to use this finding to develop one of its first investigational drugs, RLY-4008.

Over about a year and a half, Relay scientists predicted how FGFR1 and FGFR2 fold, studying whether drug candidates could bind and attach to FGFR2 without binding to FGFR1. Their process combined virtual screening—digitally testing millions of compounds against targets—with laboratory work to validate the model's predictions for a few of the most promising compounds.

In this regard, Relay is not much different from other biotech companies that use virtual compound libraries to screen for "hits" against targets. But by incorporating protein motion simulation into the process, the company hopes to do so more efficiently and more ambitiously discover new pathways where drugs might work.

"All the puzzle pieces were on the table, but no one had ever really put them together," Patel claimed.

Cell tests suggest Relay may be on the right track with RLY-4008, showing it is 200 times more potent at blocking FGFR2 than FGFR1. In comparison, Pemazyre and Balversa are much less selective.

Relay plans to begin clinical studies of RLY-4008 in cancer patients later this year.

This year, as Relay prepared for its IPO, it also disclosed two other drug candidates targeting well-known cancer targets. Like RLY-4008, information gained from protein motion models was a key component of their design, and this is why Relay believes they may be superior to other investigational drugs.

A Phase 1 clinical trial targeting a protein called SHP2 began in January.

"The toughest scientific challenge"

By choosing cancer as the first testing ground for its goals, Relay places itself in intense competition. In recent years, cancer drug development has advanced rapidly, driven by progress unrelated to protein motion.

Targeted drugs have become the standard treatment for tumors driven by multiple genetic mutations. Recently, biotech companies like Array Biopharma and Loxo Oncology have made their names by designing more effective and less toxic therapies. Genes previously considered "undruggable," such as KRAS , have also been conquered thanks to advances in medicinal chemistry.

Relay's challenge is to prove that mapping protein motion can push this rapidly advancing field even further.

Understanding how proteins move seems like an obvious advantage. But it remains unclear whether such knowledge is necessary in a world where drug discovery technologies are constantly emerging and capable of discovering and developing effective new therapies.

"Deciding which problems can be solved with this technology, I think, is the toughest scientific challenge," said Boger, who faced similar issues when creating Vertex.

Take the development of antiviral drugs that block HIV protease, an enzyme crucial for viral replication. The structure of this protein was solved in the late 1980s, paving the way for some of the earliest HIV drugs—an early success of structure-guided drug design.

HIV protease turned out to have two swingable flap structuresthat cover the site where drugs bind to the protein. How these flaps move, and how their motion affects drug candidates, is clearly important.

"You can model the motion of HIV protease beautifully," Boger said, "but the problem was already solved without doing the simulation."

Now that Relay has selected its first three drug targets, the company faces pressure to prove that the information revealed by its models enables it to create drugs that others cannot.

"Progress is happening everywhere," said Jami Rubin, a partner at investment firm PJT Partners and a Relay board member, in an interview. "We have to choose our entry points well."

The next revolution?

In 1981, eight years before Boger founded Vertex,a Fortune magazine coverdeclared that designing drugs with computers was part of "the next industrial revolution."

In the four decades since, computer models have become deeply integrated into drug discovery, with each new advance promising to solve the industry's high failure rates. First came computational molecular modeling, and now machine learning.

In seeking an edge, drug developers combine these computational tools with time-tested experimental methods. Companies like Schrӧdinger, Insitro, and Nimbus Therapeutics have been built around merging these two fields. Relay is doing the same, making protein motion its signature.

"Ten years from now, all small-molecule drug development will be done through this combination of disciplines," said Relay's Patel.

Yet, despite the enormous potential of computational drug discovery, the industry has not yet solved its core problem. Most drugs that enter clinical testing still fail, regardless of which technology facilitated their discovery or selection. Biology still trumps software, at least most of the time.

"A stumbling block for technology-based companies is that they sometimes get infatuated with their own technology because they don't understand the old days," Boger said. He hinted that Relay, with its personal connections to Vertex and other well-tested drug developers, is less likely to fall into that trap.

However, even if Relay proves its approach, early success does not necessarily lead to approved drugs or easy replication by others.

Boger knows this well. Vertex, despite its pioneering research, took two decades to obtain its first drug approval. During that time, the biotech company experienced the setbacks that are the norm in the industry, not the exception.

Boger said that if Relay has chosen the right drug targets, it may have a first-mover advantage in protein motion. But it will likely face a similar journey. "The odds are still not great," he said.