Genixcure’s AAV platform technology develops new AAV capsids capable of not only expanding the
tropism—the inherent advantage of wild-type AAVs—but also reducing clinical risks, all based on its
proprietary AI system, InsightMiner™. In addition to the enhanced tropism and safer clinical profile, it also
helps us with predicting peptide sequences for BBB penetration and/or evasion of neutralizing antibodies.
InsightMiner™ is designed to successfully overcome several limitations of traditional AAV engineering
methods that rely on random library approaches.
It is trained on human clinical-centered datasets to ensure that capsids functioning in animal models can also be translated seamlessly into human applications.
Unlike conventional methods limited to AAV variants, InsightMiner™ uses the evolution of all known viral variants as its training dataset. Therefore, the sources of inserted or substituted sequences are not restricted to AAV variants alone.
Global competitors using rational design or directed evolution typically require the direct injection of updated AAV libraries into animals, repeating this cycle dozens to hundreds of times over several years to obtain a single lead capsid. InsightMiner™, by applying the concept of “AI screening,” aims to bypass this entire process and directly identify lead capsids—ushering in a breakthrough in efficiency.
Among Genixcure’s Gen2X capsids, the CNS-targeting lead capsids are XOB031 and its variants. The key experimental data are as follows: