NORVARA THERAPEUTICS
London
Computational Virology & Gene Therapy
I. PLATFORM PHILOSOPHY
At Norvara, we approach viral vector engineering as a constrained optimization problem. Traditional gene therapy has been hindered by the inherent limitations of wild-type Adeno-associated Virus (AAV) tropism, which frequently results in hepatic sequestration and off-target toxicities.
Our research platform utilizes advanced generative AI and deep structural modeling to move beyond empirical, library-based screening. By integrating biophysics with high-dimensional latent-space modeling, we explore the vast combinatorial sequence space of AAV capsids with precision. We are currently utilizing this architecture to prioritize Renal Selective Targeting—engineering capsids that exhibit high-affinity tropism for glomerular and tubular cellular populations while maintaining a negligible profile in the liver.
II. SCIENTIFIC FOCUS
Our computational pipeline focuses on the high-fidelity refinement of three critical delivery vectors:
- Renal Tropism & Selective Targeting: We identify unique structural motifs that permit site-specific extravasation into renal tissues. By modifying the capsid surface at sub-angstrom resolution, we decouple therapeutic delivery from the default hepatic sink.
- Immunogenic Evasion: We perform predictive mapping of antigenic surface epitopes using machine learning. We generate structural variants that maximize capsid stability while minimizing cross-reactivity with existing human seroprevalent antibodies.
- Scalable Manufacturing Dynamics: A computationally "optimal" capsid is valueless if it cannot fold correctly within the production host. We apply rigid biophysical constraints during the generative phase to score sequences for thermodynamic stability, ensuring that our lead candidates maintain viable titers at industrial scale.
III. COMPUTATIONAL EXECUTION
Our research operates in a "dry lab" modality. We do not rely on high-throughput, error-prone wet-lab iterations for initial discovery. Instead, our platform utilizes a multi-stage pipeline: In-silico generative proposal, structural refinement via energy minimization, and predictive tropism validation. Only those sequences that satisfy all biophysical and immunogenic constraints are advanced to our academic partners for final biological confirmation. This workflow allows us to compress years of R&D into cycles of focused computational synthesis.