Mechanistic Model for AAV Production in HEK293 Cells (Python Port)

This repository contains a Python translation of the mechanistic model for recombinant adeno-associated virus (rAAV) production via triple transfection of HEK293 cells.

This is an open-source Python adaptation designed to run without proprietary software licenses (e.g., MATLAB). The Nguyen et al. (2021) model is based on literature derived rates and (PEI induced AAV5 transfection) experimentally derived rates. It is validated on experimental data generated by another independent study. See reference below for more detail.

Credits and Original Authors

This code is a direct port of the MATLAB model originally developed and published by the Braatz Group at MIT.

• Original Paper: Nguyen et al. (2021) - Mechanistic model for production of recombinant adeno-associated virus via triple transfection of HEK293 cells
• Original MATLAB Repository: tamntnguyen/AAV-Triple-Transfection-Mechanistic-Model

Comments

Example use:

Mechanistic modelling can help us to explain whether an observation has biological meaning or is anomalous, which helps us use resources solving problems that are worth solving rather than dead ends.

We used this model to support a hypothesis regarding the effects of a potential cell cycle inhibitor on the productivity of the process. The model accurately predicted the relative differences in productivity we observed when provided only with different cell growth profiles as input. This suggests that growth inhibition itself can mathematically account for productivity differences when considering only the components in the ODE system. In our case, this supported our conclusion that the cell cycle inhibitor allowed greater productivity by preventing the dilution of intracellular plasmids post-transfection.

Considerations:

The following limitations must be kept in mind before using: Degradation rate of plasmid complexes unknown, so media exchange at t=6 simulated to remove extracellular plasmid concentration. Rate of plasmid uptake is based on experimental results from PEI-mediated plasmid uptake. This will differ to rates from other transfection reagents. The mechanistic model itself is based on the AAV5 process and validated on AAV5 data.