Technology ID
TT-0125
Description
A new technology for cell-specific capsid engineering leverages AI-based protein design to optimize AAV capsid sequences with desirable properties. This approach integrates the library construction, screening, sequencing and deep learning of diverse AAV capsids, enabling the development of a robust AI framework. High-quality data generated from multiple animal models and human brains have been utilized to fine-tuning the pre-trained foundation model, with the ultimate goal of creating a comprehensive collection of highly efficient AAV capsids for precise and targeted gene delivery to specific cell types.
Content
Foundation model (100B) built from evolutionary databases in human virome (largest atlas of life sciences data in pre-trained model)
Experimental screenings in multiple animal species with high-quality labelled dataset generated from 500M to 1B sequence reads
Positively and negatively enriched capsids of desired characteristics, selected and validated in multiple animal models and human brain tissues
Novel capsids designed with - Shorter Production Time: 6X Faster (approx. 5-6 months) - Efficiency: 20X higher specificity (>60-70% targeting) - Low Cost & Competitive Pricing: 5X higher production
Inventors
Applicable Industries
Potential Application
Technology Focus
TRL
Images
