Preclinical findings from WhiteLab Genomics demonstrate that AI-designed viral vectors achieved approximately 50-fold higher brain DNA enrichment than AAV9 following intravenous injection in mice, while remaining undetectable in the liver.
WhiteLab Genomics has announced new preclinical data showing that viral vectors designed entirely using artificial intelligence achieved approximately 50-fold higher DNA enrichment in the brain compared to the commonly used AAV9 vector. All of this was achieved while remaining undetectable in the liver following a single intravenous injection in mice.
The findings were presented at the ASGCT Annual Meeting and were generated in collaboration with Dr. Françoise Piguet and the GENOV laboratory at the Paris Brain Institute.
The results could significantly improve genomic medicine for patients by addressing one of the sector’s most persistent challenges in delivering therapies to the brain without the need for invasive surgery.
Tackling the blood-brain barrier
Many gene therapies currently being developed for neurological diseases require direct injection into the brain or spinal fluid because the blood-brain barrier prevents most medicines from reaching the central nervous system through the bloodstream.
Even AAV9, one of the most widely used delivery vehicles in gene therapy, can lead to substantial exposure in organs such as the liver, potentially creating safety concerns and limiting dosing flexibility.
Many gene therapies currently being developed for neurological diseases require direct injection into the brain or spinal fluid
WhiteLab Genomics believes its artificial intelligence-designed vectors may provide a safer and more effective alternative by enabling therapies to reach the brain through a standard intravenous injection while reducing off-target exposure in the liver.
In wild-type mice, the company reported that its ALFRED-designed vectors demonstrated approximately 50 times greater DNA enrichment in the brain than AAV9, produced no detectable liver signal and achieved functional targeting following standard intravenous administration.
AI replaces traditional trial-and-error approaches
Most gene therapies today are developed using large-scale experimental screening methods that test hundreds of millions of viral variants, often with limited understanding of why certain vectors succeed.
WhiteLab Genomics is attempting to move away from this process through its proprietary AI platform, ALFRED, which identifies the cell-surface receptors responsible for tissue access before designing vectors around those biological mechanisms.
Rather than relying on random screening, the platform engineers vectors based on predicted biological interactions.
The company said this enables a more rational approach to vector design, supports the development of smaller and more efficient experimental libraries and could accelerate development timelines while improving translational understanding.
Industry implications
The findings highlight growing interest in the use of artificial intelligence within genomic medicine and drug development, particularly in areas where biological complexity has historically slowed progress.
“Crossing the blood-brain barrier without surgery has been one of the biggest unsolved problems in genomic medicine,” said David Del Bourgo, CEO of WhiteLab Genomics. “These results validate both our platform and our belief that AI can fundamentally change how gene therapies are designed, moving the field from brute-force screening toward rational engineering.”
Crossing the blood-brain barrier without surgery has been one of the biggest unsolved problems in genomic medicine
WhiteLab Genomics are now applying its AI-driven platform across multiple genomic medicine programmes and is continuing studies in larger animal models as it looks to advance the technology further toward the clinic.
No comments yet