Boosting pharma’s falling productivity with AI in drug discovery

Image/Generate Biomedicines
Mike Nally Generate Biomedicines ai drug discovery

The pharmaceutical industry is struggling with mounting costs of drug development. Generate Biomedicines’ CEO, Michael Nally, outlines the firm’s mission to tackle the trend by making drug discovery scalable.

Over the decades, the pharma and biopharma industry has suffered a growing stagnation in productivity. According to a report from University College London in 2018, for example, the number of new drugs approved by the U.S. Food and Drug Administration (FDA) per billion dollars spent on R&D has halved roughly every nine years since 1950.

The arrival of the COVID-19 pandemic in 2020 coincided with an improvement in productivity, especially as governments, regulators, and life sciences companies rushed to develop vaccines and treatments for COVID-19. However, drug development still faces many obstacles such as the slow, trial-and-error-based process of early drug discovery.

According to Michael Nally, CEO of the U.S. drug discovery company Generate Biomedicines, one of the factors driving this problem is that “the artisanal nature of drug discovery is inherently difficult to scale. While the unmet needs continue to be extraordinarily profound, it just costs more to develop every new drug.”

Earlier in his career, Nally spent almost 18 years at Merck (known as MSD outside of North America), eventually reaching the position of Chief Marketing Officer. During this time, Nally became acquainted with Flagship Pioneering, the venture capital (VC) company famed for founding the mRNA vaccine giant Moderna. 

“I always had great admiration for the work that was happening in the biotech sector,” said Nally, adding that the biotech sector has become “a major part of the broader pharmaceutical industry supply chain where the innovation and these new drugs are coming from these smaller entities.”

In 2021, Nally joined Flagship and led Generate Biomedicines — one of Flagship’s portfolio firms founded in 2018 — as CEO. A strong attraction for Nally was Generate’s focus on protein engineering as a source for novel protein and antibody drugs.

“The great hope for the industry is that we will understand biology in a fundamentally more profound way than ever before,” said Nally. “[Generate Biomedicines] is a truly computationally native company that sits at this intersection of science and technology. If you can harness that power, every part of the drug discovery and development process could be improved.”

Generate uses machine learning (ML) and other computational tools to design protein drugs. To fuel the calculations, the company uses a collection of protein structures in a data bank in addition to amino acid sequence data generated from the genomics revolution. This allows the team to deduce how proteins might behave based on their sequences.

With this collection of knowledge and computational power, Generate is designing protein drugs that can perform much better than current drugs at binding to their target, which could reduce the dose needed to benefit patients. Additionally, Generate is developing protein drugs and antibodies that can avoid recognition by the patient’s immune system, which often attacks and blunts the effect of these therapies.

“On the other side of that, you can make things more immunogenic,” added Nally. “For vaccines, a more robust immune response is actually the desired parameter. There’s a whole host of developability and manufacturability considerations that we measure in each individual case.”

Many biotech companies are employing artificial intelligence (AI) and ML to speed up drug discovery. One notable example is Exscientia, which focuses on small molecule drug discovery in the U.K. Another is Deep Mind, whose AI-driven AlphaFold tool has made big strides in the prediction of protein structures based on amino acid sequences. According to Nally, Generate stands out with its focus on developing brand new proteins and other large molecule drugs.

“The work that AlphaFold has done is truly monumental for the field,” said Nally. He added that Deep Mind’s work centers on predicting existing protein structures that have already been studied and can easily be validated. In contrast, Generate focuses on designing new, unmapped protein structures that have never before existed in nature. 

Generate is applying its technology to the preclinical development of 14 candidate drugs in the fields of oncology, immunology, and infectious disease. Five of these candidates are being co-developed with Amgen as part of a collaboration deal worth up to $1.9 billion. To finance its business and drug development, Generate bagged $370 million in a Series B financing in late 2021.

Generate’s mission is to solve the pharma industry’s productivity crisis by making drug discovery more scalable than it is at present.

“Whenever complex fields have become engineerable, industrial revolutions occur,” said Nally. “This is an era of generative biology that will allow us to make the field increasingly engineerable and therefore address many of the productivity challenges. Couple that with the fact that when you put the computer at the center of the drug discovery engine, computational power is inherently scalable.”

It’s tough to know when biology as a whole will become engineerable, but Nally sees various aspects of the field making this transition within a decade. For now, Generate is putting its bets on protein engineering.

“Proteins drive 95% of biological processes,” Nally explained. “If you can master that domain, in many respects, you have the building blocks of a lot of underlying biology.”

Newsletter Signup - Under Article / In Page

"*" indicates required fields

Subscribe to our newsletter to get the latest biotech news!

This field is for validation purposes and should be left unchanged.

Suggested Articles

Show More