Rare diseases are illnesses that affect 1 in 2,000 individuals. Often described as chronic, they are long-lasting and serious, causing disabilities and, in many cases, even leading to early death.
With over 7,000 known conditions, rare diseases collectively impact 300 million people globally, demanding urgency in intervention and care. For decades, the core challenge has been a scarcity of patient data, a “data desert” stalling the development of new therapies.
A new wave of progress is here, driven by technology. While investment and research have always been present, they’ve been hampered by the lack of data. AI is transforming this landscape, turning data scarcity into an opportunity. By analysing fragmented datasets, AI is accelerating research and attracting new capital to this historically underfunded field.
From data desert to digital oasis
In rare diseases, patient pools are small, scattered, and often underdiagnosed. Traditional research models are not always up to scratch when patient pools are scarce. The data itself is a complex mix, from genomic sequences to real-time feeds from wearables.
For example, generative AI can now create synthetic datasets, giving researchers model diseases and test hypotheses without compromising privacy. Beyond creating data, AI excels at finding the needle in the haystack: subtle biological patterns invisible to the human eye. It changes the equation entirely, analysing complex datasets typically handled by hundreds of researchers. Researchers at Carnegie Mellon proved this with KGWAS, a deep-learning method that uncovered twice as many meaningful genetic links for certain conditions, all from smaller patient groups.
Capitalising on speed
The most profound impact of AI is on speed. In drug discovery, years of research can now be compressed into months. By simulating how therapies will interact with cells before a single patient is enrolled in a trial, companies can “fail faster,” focusing capital on the most promising candidates. This isn’t just a marginal improvement; it’s a shift that could boost the success rate of therapies entering clinical trials from an estimated 60% to over 80%.
Recent applications of large language models in analysing single-cell data demonstrate how vast datasets that once took months to process can now be handled in a fraction of the time. This acceleration is tangible. Scientists now realistically expect to see breakthroughs within their own lifetimes, a prospect that was pure science fiction just a decade ago. This newfound speed is supported by major advances in computing, with modern GPUs providing the raw power to model human biology with precision.
Precision AI and governance
For rare diseases, where each case is unique, a one-size-fits-all approach is ineffective. Precision AI, which leverages vast genomic and clinical data, creates opportunities to tailor therapies to a patient’s specific genetic makeup, lifestyle, and environment, moving us from general treatment to targeted, personal medicine.
Powerful tools are only part of the story, however. The foundational rule of data science — garbage in, garbage out — is especially true in medicine. This demands a robust infrastructure built on high-performance computing, secure storage, and meticulous data governance.
Progress also requires clear regulatory frameworks, much like the safety standards that govern the automotive industry. Singapore’s model AI governance framework, launched last year, is a promising template for how to innovate responsibly. With this foundation, concepts like patient “digital twins” are moving from theory to reality. We can test a therapy on a virtual model of a patient first, personalising medicine and reducing trial and error before it ever reaches a human being.
A call for recognition and collaboration
The burden of rare diseases is immense, affecting not only patients but also their caregivers, families, and healthcare systems. The social and psychosocial strain, including isolation and discrimination, are compounded by the high economic costs of diagnosis, treatment, and productivity loss. The individual rarity of these diseases creates a gap in research and drug development that affects available therapies.
We are finally entering an era where AI’s speed and scale can match the urgency of patients’ needs, offering a solution to these burdens. It is through broad ecosystem collaboration among innovators in technology, science, and government that a credible and accelerating path toward a cure for the millions living with a rare disease can be built.
This is a collaborative journey that requires us to ensure these AI-driven breakthroughs translate into a healthier, more humane future for all.
