MSD, which has long dominated the global cancer market with its blockbuster immuno-oncology therapy Keytruda, is investing heavily in artificial intelligence (AI)-based protein design. The strategic goal is to build the robust experimental datasets required for AI model training, enabling researchers to identify novel drug candidates with unprecedented speed and precision.
Under a newly announced joint research and licensing agreement, MSD will partner with AI-driven drug-design specialist Protillion Biosciences in a deal valued at up to $510 million (approximately 770 billion won). Protillion will receive an undisclosed upfront payment and remains eligible for subsequent milestone payments tied to research, development, and commercialization success.
At the heart of the partnership is Protillion’s proprietary protein engineering platform, Prot-MaP. The system is capable of generating large-scale experimental datasets that protein-design AI can learn from, while quantitatively analyzing the characteristics of a vast array of protein candidates. MSD plans to pair this platform with its own in-house drug discovery capabilities to identify multiple new therapeutic targets.
Why AI Drug Discovery Hinges on 'Good Data'
Leveraging AI in modern drug development requires far more than raw computing power; success depends entirely on the availability of high-quality experimental data. If training data is scarce or skewed, AI systems risk generating plausible-sounding predictions for molecular candidates that ultimately fail to prove viable in real-world clinical development.
Protillion’s Prot-MaP platform is engineered to mitigate these exact bottlenecks. By synthesizing the precise training data needed for protein-design AI, the platform allows researchers to efficiently screen, compare, and analyze the characteristics of massive pools of protein candidates simultaneously.
This type of advanced protein design is critical for developing complex biologics, including antibody therapies, bispecific antibodies, and antibody-drug conjugates (ADCs). The clinical viability of these treatments depends heavily on how precisely a therapy binds to a disease-related target, how stably it functions within the human body, and its ability to minimize off-target binding.
According to Protillion, its technology can isolate sophisticated biological candidates that remain elusive via conventional methods—such as antibodies that alter their binding properties under specific acidity conditions, or protein candidates engineered to hit multiple disease targets at once. Specific disease targets and candidate molecule names remain undisclosed under the current agreement.
The Global Race for Next-Gen Biologics
Global pharmaceutical companies are aggressively accelerating AI- and data-driven drug discovery as competition intensifies to secure next-generation oncology pipelines.
This trend is strongly mirrored in markets like South Korea, where a shifting cancer burden continues to drive demand for precision therapies capable of overcoming the limitations of first-generation treatments. According to the country's 2023 National Cancer Registration Statistics, thyroid cancer remains the most prevalent, followed closely by lung, colorectal, breast, stomach, prostate, and liver cancers. As patient demographics evolve and treatment strategies become highly segmented by tumor subtype, the clinical necessity for targeted biologics continues to surge.
While MSD maintains a dominant position in the immuno-oncology market through Keytruda, the company is moving proactively to build out its portfolio ahead of future patent expiries. The Protillion agreement follows a series of high-profile collaborations by MSD aimed at integrating AI, data analytics, and advanced antibody discovery. In March, MSD inked a $1 billion cloud computing partnership with Google Cloud (worth roughly 1.5 trillion won), alongside a $2.2 billion inflammatory bowel disease discovery deal with Quotient Therapeutics (around 3 trillion won), and a separate antibody discovery pact with Infinimmune.
Ultimately, the Protillion deal underscores a pragmatic industry shift: rather than treating AI as a total replacement for human researchers, leading drugmakers are focusing capital on the underlying experimental data and analytical infrastructure required to optimize discovery. For MSD, as it charts its multi-billion-dollar path beyond Keytruda, the ability to generate, refine, and interpret high-fidelity data is rapidly becoming the ultimate competitive battleground in biologics development.
