Novo Nordisk Expands AI-Driven Drug Discovery Partnership with Valo Health in $4.6 Billion Deal

On January 8, 2025, it was reported that Novo Nordisk announced an expanded agreement to leverage Valo Health's (hereinafter referred to as Valo) extensive human datasets and AI-driven computing power to discover and develop novel treatments for obesity, type 2 diabetes, and cardiovascular diseases.

Under the terms of the expanded agreement, Valo is entitled to receive an upfront payment, equity investment, and potential near-term milestone payments totaling $190 million, and is now eligible for milestone payments on up to 20 drug programs, with nine additional drug programs, along with research and development funding and potential royalty payments, amounting to approximately $4.6 billion in total.

In September 2023, Novo Nordisk and Valo Health reached a collaboration agreement with a $60 million upfront payment and a total value of $2.7 billion. The agreement allows for the development of up to 11 drug projects, primarily focused on cardiovascular diseases. Valo Health's Opal Computational Platform (hereinafter referred to as Opal) can identify and validate novel disease targets and molecules that can act on these targets. Additionally, Opal is capable of performing simulations to predict the safety and efficacy of compounds. Novo Nordisk hopes that Opal will complement its existing technology to develop new drugs in the cardiovascular field.

At that time, Novo Nordisk's product portfolio had already expanded from diabetes to obesity treatment and was preparing to extend into other cardiometabolic diseases. Earlier, in September 2022, Novo Nordisk secured the rights to VENT-01, an NLRP3 inhibitor developed by Ventus Pharmaceuticals, with a $70 million upfront payment and a commitment of up to $700 million. VENT-01 is a small molecule with potential to treat liver, kidney, and cardiometabolic diseases.

Before partnering with Valo, Novo Nordisk also acquired lnversago Pharma and Embark Biotech for $1.1 billion and $500 million, respectively, each offering distinct therapeutic approaches to cardiometabolic diseases.

Valo, incubated by Flagship, a well-known U.S. biopharmaceutical venture capital firm, was established in 2019 and officially came out of stealth mode in 2020. It leverages data and AI to develop the next generation of drug discovery platforms, with a pipeline covering 17 projects across therapeutic areas such as cardiovascular/metabolic/renal diseases, cancer, and neurodegenerative diseases.

Since its establishment over the years, Valo Health's total financing has exceeded 450 million US dollars, with the highest round being a 300 million US dollar Series B financing. In 2021, when companies like Schrödinger, Relay, Recursion, and Exscientia went public with soaring valuations, Valo Health, which was only at the Series B stage, planned to go public through a SPAC merger with a well-known Silicon Valley venture capital firm. However, the company subsequently announced the termination of its IPO process.

It is worth mentioning that Valo is a fully integrated small-molecule drug discovery and development company, covering target identification, molecular discovery, and clinical development. Its core product, the Opal platform, is based on longitudinal datasets from thousands of sources. It not only predicts molecules that can be used as new drugs but also forecasts how these drugs will affect patients' bodies, effectively avoiding side effects of new drugs during the R&D process.

Behind the capabilities of the Opal platform lies a vast accumulation of underlying data. Between 2019 and 2020, Valo acquired AI drug discovery company Numerate and biopharmaceutical firm FORMA Therapeutics, gaining assets such as two early discovery laboratories and R&D libraries. Numerate’s platform includes over 30,000 models, 70 trillion molecules, and more than 25 drug projects. These assets have further strengthened Valo's Opal platform.

In addition, Valo has also partnered with genomics company G3 Therapeutics to access G3 Therapeutics' aerobic metabolism database, which aids Valo in discovering disease-related biomarkers and targets, thereby accelerating the drug development process.

Besides the data scale, another highlight of the Opal platform is its choice to incorporate human data rather than mouse cell data, which has been commonly used in the industry. Undoubtedly, human data can make drug development faster and more efficient while also reducing costs. For Valo, the application of human data can become a core competitive advantage for its algorithm platform.

In addition, the end-to-end design of the Opal platform has significantly accelerated the speed of drug development. At the same time, with data utilizing a single integrated architecture, data and information can be shared throughout the entire drug research and development process, providing new methods for the development of disease treatments. Compared with traditional methods, the Opal computing platform can perform computational and empirical screening of trillions of molecules within weeks, thereby identifying new target areas for development.

However, despite having powerful underlying data and algorithm logic, the past year has not been smooth sailing for Valo.

On the last day of 2024, Valo Health announced the topline data from its Phase 2 SPECTRA study of OPL-0401 in patients with diabetic retinopathy (DR). The results showed that the study did not meet its primary or secondary endpoints, and development has been suspended. Industry insiders generally believe that the failure of this Phase 2 result was not without precedent, as the company had already gone through two CEOs before the Phase 2 data was released.

Fortunately, OPL-0401 was not discovered through Valo's AI discovery platform but acquired from Sanofi. However, Valo has not provided any specifics on when or in what therapeutic area it might advance a candidate drug using its Opal platform.

Novo Nordisk’s Chief Scientific Officer stated that the company is very satisfied with the progress of its collaboration with Valo and looks forward to further strengthening its R&D capabilities in chronic diseases, particularly obesity and type 2 diabetes, through an expanded partnership. However, by the end of 2024, due to disappointing data from the highly anticipated "next-generation weight loss drug," Novo Nordisk will need to seek new growth drivers.

Over the past year, Novo Nordisk's business positioning has remained relatively concentrated, with its products primarily covering four major therapeutic areas: diabetes, obesity, rare diseases (including blood disorders and growth disorders), and other chronic conditions (such as atherosclerotic cardiovascular disease, heart failure, non-alcoholic fatty liver disease, and Alzheimer's disease).

Novo Nordisk's strategy in the chronic disease field combines internal growth with external expansion, while actively venturing into emerging therapeutic areas. The global development of GLP-1 analog indications has expanded from type 2 diabetes and obesity to include non-alcoholic fatty liver disease, pediatric and adult type 2 diabetes, Parkinson’s disease, and the improvement of diabetic complications, among others. Novo Nordisk's financial reports indicate that, aside from diabetes and obesity indications, semaglutide has also been deployed across cardiovascular diseases, MASH, CKD, Alzheimer’s disease, heart failure, and other areas, involving various mechanisms of action and development platforms—all of which represent clinical fields with immense potential.

It is worth mentioning that the field Novo Nordisk is focusing on also includes cardiometabolic disease treatment, which is in a "renaissance phase." To gain an early advantage in this field, the assistance of AI is considered essential.

Taking the field of heart failure as an example, there is an urgent need for breakthroughs in treatment drugs, and new blockbuster varieties with proven efficacy are rapidly gaining traction: The guidelines for heart failure treatment have evolved from the traditional "gold triangle" (ACEi/ARB+BB+MRA) to a new "gold triangle" (ACEi/ARB/ARNI +BB+MRA), and then to the new quadruple therapy (ARNI or ACEi/ARB+SGLT2i+BB+MRA). Antidiabetic drugs have also demonstrated cardiovascular benefits and have been newly incorporated into the quadruple therapy as foundational treatments for HFrEF, significantly enhancing their market potential.

In the long term, there are still many unmet needs in the treatment of cardiometabolic diseases, and the development of new products combined with the expansion of indications will continue to drive market growth. In addition, improving the prognosis of related diseases has always been a direction for new drug development. Currently, major multinational corporations (MNCs) are applying cutting-edge technologies such as stem cell therapy and gene therapy in the field of heart failure, as well as intensifying efforts in exploring new targets.

The key question now is whether Valo and other AI-driven drug discovery companies can improve the success rate of drug development remains to be tested.

For the few successful pipelines, the contribution of AI is limited. For instance, Takeda's $6 billion acquisition of the TYK2 inhibitor TAK-279 was once highly regarded as an "AI-driven drug discovery," but in reality, it was developed by Schrödinger and Nimbus referencing a molecular structure published by BMS and optimized through FEP (Free Energy Perturbation) calculations.

Another "criticized" aspect of AI pharmaceuticals is the poor performance of several Biotech companies, which were listed at high prices a few years ago, in the capital market. Looking back at 2024, AI drugs failed one after another, and most of them are still in the critical Phase 2 clinical trials. In September 2024, Recursion announced limited efficacy in the Phase 2 data for its rare disease small molecule drug REC-994; in December, Bioage Labs' weight loss drug Azelaprag showed elevated transaminase levels in some participants, forcing the company to halt its Phase 2 clinical study.

An Increasing Number of People Realize That There Is a Gap Between the Promises Made by Many AI Pharmaceutical Companies and the Actual Situation.

In addition, early AI pharmaceutical companies relied too heavily on data-driven approaches to accelerate drug discovery. However, in situations where data is scarce, this method may generate compounds similar to known drugs or repeat previously failed ones, which once made AI drug discovery disappointing. Now, AI pharmaceutical companies are adjusting their strategies, shifting toward personalized treatment or addressing more fundamental issues like drug target selection, rather than focusing solely on drug optimization.

However, the collaboration between AI pharmaceuticals and MNCs can still bring new possibilities.

MNCs have abundant funding and data, extensive experience in drug development, and usually prefer to use tools developed by professional companies. For AI pharmaceutical companies, cooperation with MNCs is often seen as an opportunity to gain necessary resources, accelerate the drug development process, and expand their business scope. After all, AI pharmaceutical companies tend to specialize deeply in a specific technology or disease area. If an AI pharmaceutical company can address the pain points of MNCs with unique technology and help them understand the technical advantages of its AI drug discovery platform, it will often attract more investment from MNCs.

Moreover, cost reduction is almost a key focus for all pharmaceutical companies at present. Multinational corporations (MNCs) generally have high expectations for the application of AI in drug development, viewing it as an essential tool to improve efficiency, cut costs, and accelerate the time to market for new drugs. They are often unwilling to miss collaboration opportunities and are willing to place greater hopes on AI platforms.

As Valo's CEO Brian Alexander pointed out in the announcement: OPL-0401 did not originate from Valo's discovery platform, so the failure of the pipeline does not represent the failure of the platform.