Accelerating rare disease cures with ASOs, gene editing, and AI

What It Covers

Professor Matthew Wood of Oxford's Harrington Rare Disease Centre outlines how antisense oligonucleotides, gene editing, and AI tools can be developed as scalable platforms to accelerate therapies for the estimated 500 million rare disease patients worldwide, with a target of delivering treatments within 100 days of genetic diagnosis.

Key Questions Answered

• •ASO Platform Scalability: Antisense oligonucleotides share identical chemistry across applications — only the gene-targeting sequence changes. With nearly 20 approved drugs already generating safety data, researchers and regulators can now treat ASOs as a reusable platform, eliminating redundant toxicology studies worth millions of dollars and years of development time per new indication.
• •Regulatory Reform Pathway: Current drug approval frameworks were designed 60 years ago for small-molecule drugs and large patient populations — the opposite of precision genetic medicines. The UK's MHRA is actively exploring platform-based regulation that allows new ASOs targeting different genes to leverage existing safety datasets rather than restarting full preclinical programs from scratch.
• •AI-Accelerated Drug Identification: Professor Wood's team built Oligo AI, trained on patent database records rather than proprietary pharma data, to fast-track identification of potent lead oligonucleotide molecules. The next development phase targets safety prediction. This approach bypasses the hundreds of millions of dollars needed to generate equivalent experimental screening datasets independently.
• •Newborn Screening as Near-Cure Strategy: Treating rare disease patients presymptomatically at birth, rather than after clinical decline, transforms marginally effective drugs into near-curative interventions. Combining approved ASO therapies with national genome newborn screening programs — currently in pilot phase in the UK — could prevent disease expression entirely rather than slowing progression in already-symptomatic children.
• •Gene Editing Delivery Gap: CRISPR-based gene editing remains unsuitable as a brain disease platform today because no viable delivery mechanism exists for the central nervous system. Since 60–70% of rare diseases affect the brain, solving CNS delivery is the critical bottleneck. Progress is more advanced in liver and blood, where delivery barriers are substantially lower.