Parkinson’s disease remains one of the most challenging neurodegenerative disorders to treat, in part because of its biological complexity and clinical variability. More than 10 million people worldwide are estimated to be living with Parkinson’s, and current therapies largely focus on symptom management rather than altering disease progression. Despite decades of research, disease-modifying treatments have yet to reach patients.
Against that backdrop, Valo Health has announced a partnership with Merck KGaA, Darmstadt, Germany, aimed at discovering and developing novel treatments for Parkinson’s disease and related disorders. The partnership includes upfront and potential milestone payments totaling more than $3 billion, along with royalties and R&D funding, underscoring the scale of investment flowing into AI-enabled approaches for neurodegenerative disease.
The collaboration will combine Merck KGaA’s neurology research expertise with Valo’s AI-enabled human causal biology and closed-loop discovery platforms to identify and validate new disease targets and rapidly generate preclinical small-molecule candidates.
“Our research engine is focused on delivering meaningful medicines for patients with high unmet medical needs,” said Amy Kao, Global Head of Neurology & Immunology Research Unit at Merck KGaA, in the company’s announcement. She added that Valo’s platforms “will help sharpen target selection and streamline drug discovery, enabling us to advance the most promising candidates faster.”
Using human data to address disease heterogeneity
One of the central scientific challenges in Parkinson’s is its heterogeneity. Patients can vary widely in symptom profile, progression rate, and underlying biology, complicating both drug development and clinical trial design. Many late-stage failures in neurodegenerative disease have been attributed to insufficient target validation or lack of patient stratification, a pattern seen across Alzheimer’s, ALS, and Parkinson’s programs.
“Despite progress in science and medicine, our understanding of this condition is largely based on limited observational data leading to broad definitions of the disease, and few treatment options that primarily manage symptoms,” Brian Alexander, CEO of Valo Health, told DDN.
Valo’s approach centers on human-derived data to address those limitations earlier in discovery. The company has access to more than 17 million de-identified patient records, including roughly 900,000 individuals with neurodegenerative conditions and about 160,000 with Parkinson’s disease specifically. These longitudinal datasets span decades in some cases, offering insight into disease progression, comorbidities, and treatment response.
“By combining genomic, transcriptomic, proteomic, and real-world clinical data, we can build causal models that link molecular changes to patient outcomes,” Alexander told DDN. These models help the biotech prioritize targets with strong human evidence and validate them early “ reducing the risk of pursuing non-causal biology,” Alexander said.
Patient stratification is a key part of that strategy. By grouping patients based on biomarkers, genetic risk factors, or progression patterns, developers may be able to design more targeted therapies and more informative clinical trials — an area of growing interest across the neurodegeneration field.
Linking target discovery to small molecule development
Once targets are identified, the collaboration will rely on Valo’s closed-loop discovery platform to move programs through hit identification and lead optimization. Unlike traditional linear workflows, the system is designed to continuously integrate AI modeling with experimental data generation.
“Similar to Valo’s Human Causal Biology platform, Valo’s closed loop discovery platform aims to align chemistry to success in patients,” Alexander told DDN. “The small molecule platform natively integrates AI modeling and experimental data generation to identify the most advanceable chemistries. In hit and early lead identification, it allows us to navigate vast spaces of accessible chemistry to deliver both series and predictive models able to drive medicinal chemistry. As programs advance into lead optimization, the iterative cycles are guided by real-time data to simultaneously improve the models and compounds, accelerating the path to high-quality candidates.”
The partnership also reflects broader industry efforts to move beyond dopamine replacement therapies such as levodopa. While symptomatic treatments remain foundational, many companies are pursuing disease-modifying strategies targeting pathways including alpha-synuclein aggregation, mitochondrial dysfunction, lysosomal biology, and neuroinflammation. Antibodies, gene therapies, and small molecules aimed at these mechanisms are currently in clinical development, though none have yet shown definitive disease-modifying benefit.
Valo’s collaboration with Merck KGaA builds on its existing partnerships, including work with Novo Nordisk in cardiometabolic disease, and underlines the broader applicability of its platform across therapeutic areas. While the disease focus differs, both programs rely on the same underlying capabilities: large-scale human data integration, causal inference, and closed-loop discovery.
As developers continue to grapple with high attrition rates in neurodegenerative disease, approaches that emphasize human biology earlier in the pipeline may play a growing role in shaping future Parkinson’s programs.
Whether these strategies can translate into durable clinical success remains to be seen, but the size and scope of the deal signal sustained industry interest in new ways of tackling one of neurology’s most persistent challenges.
