VectorBuilder and Lir Therapeutics have entered a collaboration to optimise AAV capsids using artificial intelligence, laboratory testing and manufacturing assessment.

AI-guided AAV capsid optimisation

VectorBuilder will combine its expertise in AAV discovery, capsid engineering, process development and GMP manufacturing with Lir Therapeutics’ nAAVigator artificial intelligence platform.

The collaboration will focus on selected capsids identified through VectorBuilder’s DeepCap platform, with Lir applying its AI-powered engineering and laboratory validation approach to further optimise their properties.

Rather than developing entirely new capsids, the companies will work on existing candidates with biological potential and assess whether their characteristics can be improved across biological performance and manufacturability.

The partners said the collaboration will initially evaluate multiple lead capsids across selected therapeutic applications using iterative computational engineering, experimental validation and manufacturing assessment.

Bruce Lahn, chief scientist of VectorBuilder, said: “Gene therapy doesn’t just need better capsids; it needs capsids that can make it all the way to patients.”

He added that combining VectorBuilder’s capsid engineering and development expertise with Lir’s AI platform could help address manufacturing and development challenges earlier in the AAV development process.

Focus on manufacturing as well as biological performance

AAV capsids are an important component of many gene therapy approaches because they determine how effectively a vector can deliver genetic material to target cells and tissues.

However, capsids with desirable biological characteristics can present manufacturing challenges as programmes progress towards clinical development.

The collaboration will therefore assess characteristics including manufacturing productivity, process robustness, full capsid yield and product quality alongside biological performance.

The companies also aim to investigate capsid characteristics linked to tissue specificity, off-target delivery, immune recognition and therapeutic dose requirements.

Killian Hanlon, CEO of Lir Therapeutics, said: “AI only creates value if it helps solve real biological problems.”

Hanlon added that Lir’s platform is intended to address several challenges in viral vector engineering, including dose reduction, tissue targeting and immune evasion.

The collaboration is intended to bring manufacturing considerations into the capsid engineering process earlier rather than assessing scalability only after a candidate has been selected.

Combining computational and experimental development

Lir’s approach combines computational protein engineering with experimental validation, while VectorBuilder provides capabilities spanning AAV discovery, engineering, process development and GMP manufacturing.

The companies said data generated through the collaboration will be used to support further optimisation of VectorBuilder’s capsid portfolio and could create opportunities for future therapeutic partnerships.

The initial work remains at the development stage, with no clinical programme announced as part of the collaboration.

The partnership reflects the increasing use of computational methods in viral vector engineering, while retaining laboratory and manufacturing testing as part of the development process.