I-X Technical Tutorials: Bayesian Kinetics

This tutorial introduces Bayesian statistical modelling using PyMC for analysing enzyme kinetics in bioprocess development.

🚀 Overview

You will explore how to:

• Simulate kinetic data with latent effects (e.g. temperature, pH)
• Build Bayesian models to fit kinetic curves
• Extend models with Gaussian process (GP) priors for hybrid modelling

---

📁 Tutorial Structure

Notebook	Title	Description
0_data_generation.ipynb	[Hidden] Data generation	⚠️ Not for initial use. This notebook simulates the kinetic dataset. Use it only if you want to inspect or regenerate the synthetic data.
1_bayesian_kinetics_intro.ipynb	Basic Bayesian model with mechanistic equation	🧪 Your starting point. Fit a simple model to the observed kinetic data using PyMC.
2.1_bayesian_kinetics_gp_exercise.ipynb	Hybrid modelling with latent GP (Exercise)	🔬 Advanced model using a GP prior to capture latent dependencies (temperature/pH effects).
2.2_SOLUTION_bayesian_kinetics_gp_exercise.ipynb.	Hybrid modelling with latent GP (Solution)	💡 The full solution in case you need help.

---

✅ Start Here

Start with notebook 1:

Then explore hybrid modeling with GPs:

---

📦 Requirements

The easiest way to use the notebooks is via Google Colab.
If running locally, install the dependencies via:

conda create -c conda-forge -n pymc_env "pymc>=5"
conda activate pymc_env
pip install pandas matplotlib