Limitations of Data Scale and Fundamental Errors in AI

AI models for oncology treatment require vast amounts of data for drug repurposing, but simply scaling up the data has clear limitations. Preclinical and clinical data are often scattered across multiple systems, increasing the risk of AI misinterpreting spurious correlations as causal relationships. AI cannot verify logical errors in research design or determine factual accuracy, so if the underlying data is flawed, the reliability of predictions will inevitably be compromised. Ultimately, if structured data integration does not precede the use of AI, the probability of success for clinical candidates identified by AI will be reduced.

The Need for Expert-Curated Databases

To overcome this, global bio company QIAGEN N.V. (QGEN) emphasizes high-quality data curation by experts. Its subsidiary, QIAGEN Digital Insights (QDI), provides exclusive access to reliable genomic data through the Human Somatic Mutation Database (HSMD) and the Catalogue of Somatic Mutations in Cancer (COSMIC). Genomic pathways and drug-target interaction information are refined into a normalized knowledge graph, which improves AI's learning speed and accuracy. Reliable data enables validated AI models to significantly reduce the risk of new drug development and enable meaningful expansion of indications.

Real-World Success Stories of AI-Driven Drug Repurposing

AI-based drug repurposing has already demonstrated groundbreaking results in the market. BenevolentAI used its proprietary AI platform to identify Olumiant (baricitinib), a JAK1/JAK2 inhibitor from Eli Lilly (LLY) and Incyte (INCY), as a treatment for COVID-19. This drug received FDA Emergency Use Authorization (EUA) on November 19, 2020, and full approval on May 10, 2022, contributing over $800 million in revenue to Eli Lilly during the pandemic. This is a significant case that demonstrates the reduction in development time and the market entry advantage achieved through AI, even among competing drugs such as Pfizer's (PFE) Paxlovid.

Deal Mechanics and Value Analysis of AI-Driven New Drug Development

The independent data competitiveness of biotech companies is proving its value by leading to major technology transfer agreements. Sanofi (SNY)'s 2022 partnership with Exscientia (EXAI) to identify 15 anti-cancer targets is a prime example. The terms of the deal included an upfront payment of $100 million and milestone payments of up to $5.2 billion, plus royalties. Platforms with refined data can shorten the time required for initial validation of candidate compounds by several years, enabling them to command high valuations from major pharmaceutical companies.