Bioprocess Modeling Made Easy With Hybrax

Hybrax enables hybrid neural-ODE bioprocess models to be trained with performance comparable to original literature models, demonstrated on real-world Escherichia coli and synthetic HEK293 data.

• Why it matters: Bioprocess hybrid ODE models are hard to reproduce because researchers must build bespoke data pipelines, balance-equation bookkeeping, and cross-validation training code, and there is no standard data format.
• What they did: Hybrax is an open-source Python framework that pairs a flexible bioprocess data format with a customizable end-to-end training pipeline, automatically deriving transport terms and packaging training checkpoints with code and configuration.
• The result: Re-implementations of two published models—one trained on real-world Escherichia coli data and one on synthetic HEK293 data—achieve performance comparable to the original models, supporting reuse and retraining from checkpoints.

AI summary of the abstract · may contain errors · not medical advice · read the paper.