Rather than using artificial intelligence to design the right drug for a known disease, a new wave of AI-driven platforms is being trained to pinpoint the underlying biological triggers of disease itself, potentially revealing drug targets that have never been explored before. This represents a fundamental shift in how the pharmaceutical industry approaches drug discovery, moving from treating symptoms to understanding root causes .

Why Is Finding Disease Targets Harder Than Finding Drugs?

For decades, drug discovery has followed a predictable path: scientists identify a known disease target, then search for molecules that can hit that target. AI has already proven effective at this task, with over two dozen AI-designed drug candidates currently in clinical trials . But finding the targets themselves is a different challenge entirely. "It's about finding new targets altogether, new biology, new mechanisms for solving disease," explained Nick Naclerio, founding partner at Illumina Ventures. "That's a much harder problem."

The difficulty lies in the sheer complexity of human biology. AI systems are trained on existing scientific literature and knowledge, but our understanding of how the body works remains incomplete. To train AI models that can discover entirely new disease mechanisms, researchers need massive datasets that capture what happens when thousands of genes are turned on and off across different cell types and tissues .

"AI is trained on what's in the literature, but there's a finite understanding of how human biology works," said Nick Naclerio, founding partner at Illumina Ventures.

Nick Naclerio, Founding Partner at Illumina Ventures

What Is the Billion Cell Atlas, and How Does It Work?

To tackle this challenge, Illumina announced a major research consortium in 2025 called the Billion Cell Atlas, designed to create a comprehensive map of human disease biology. The project brings together industry leaders including Merck & Co., Eli Lilly, and AstraZeneca to build what amounts to a foundational dataset for training next-generation AI models .

The Atlas will develop over 200 cell lines relevant to disease areas such as cancer, cardiometabolic conditions, and neurological disorders. Researchers will then use CRISPR gene-editing technology to systematically test what happens to billions of individual cells when thousands of different genes are switched on and off. The scale is enormous: with roughly 20,000 genes to test across hundreds of tissues, the resulting dataset will be orders of magnitude larger than anything currently available .

Once this foundational data is generated, AI models can be trained to identify patterns that reveal which genes drive disease and which biological pathways could be targeted with new drugs. This approach could uncover disease mechanisms that were previously invisible to researchers .

How Are Other Organizations Approaching Disease Target Discovery?

Illumina's effort is not the only initiative pushing this frontier. Academic researchers at Harvard Medical School have developed an AI-driven tool that can identify multiple disease drivers simultaneously and then determine which genes could address them. This approach could steer drug development toward more effective targets than traditional methods, which test one protein or drug at a time .

Other biotechnology companies are also advancing the field. Australia-based Omnigeniq developed a platform that can visualize proteins in their ligand-ready state, helping researchers understand the structural changes that lead to disease progression .

Steps to Leverage AI for Disease Target Discovery in Your Organization

• Build or Access Large Biological Datasets: The competition is now centered on assembling comprehensive datasets that can train AI models effectively. Organizations need genomic, transcriptomic, proteomic, and clinical data at scale to enable meaningful AI analysis.
• Invest in Multidisciplinary Teams: Successfully implementing AI for disease target discovery requires experts who understand data science, genomics, computational biology, and model validation. The shortage of such talent remains a significant barrier to adoption.
• Adopt Cloud-Based Bioinformatics Platforms: As biological datasets grow in complexity, cloud-based infrastructure becomes essential for managing and analyzing multi-omics data efficiently across research teams.
• Collaborate on Foundational Research: Joining consortiums like the Billion Cell Atlas or similar initiatives allows organizations to share the costs and benefits of building large-scale datasets that benefit the entire industry.

What Does This Mean for the Future of Drug Development?

The shift toward AI-driven disease target discovery is part of a broader evolution in how pharmaceutical companies approach R&D. The global AI in bioinformatics market was valued at USD 1.06 billion in 2025 and is projected to grow to USD 4.80 billion by 2034, growing at a compound annual rate of 18.63% . This expansion is driven by increasing volumes of sequencing and multi-omics data, growing applications of AI in drug discovery and precision medicine, and rising demand for cloud-based platforms capable of handling large-scale biological datasets .

One key trend accelerating this growth is the shift toward precision medicine. Healthcare and life sciences organizations are moving away from generic treatment approaches toward data-informed, personalized care. This requires AI tools capable of analyzing genomic, transcriptomic, clinical, and other biological data collectively to identify disease patterns and treatment responses more accurately . In April 2025, Illumina and Tempus announced a collaboration combining Illumina's AI technologies with Tempus's multimodal data platform to accelerate clinical adoption of molecular testing and advance precision medicine .

However, significant challenges remain. The restricted availability of skilled personnel with expertise in AI, biology, and bioinformatics poses a major limitation. Many organizations can access AI tools but struggle to implement them effectively without trained bioinformaticians and AI experts. This gap is particularly acute in emerging markets and academic research environments, where retaining specialized talent is difficult . Additionally, integrating complex data from multiple sources, formats, and scales remains a technical hurdle that organizations must overcome .

Despite these obstacles, industry experts remain optimistic about the timeline. "The next few years will be a race to build foundational models that can make leaps in generating new insights into understanding biology," Naclerio noted. "Over time, we are going to bend the curve of drug discovery productivity and see huge improvements" .

The convergence of larger datasets, more powerful AI tools, and increased biotech investment is creating a moment where the industry can finally tackle one of its hardest problems: not just finding the right drug, but understanding the fundamental biology that makes disease possible in the first place.