ENKIE - The ENzyme KInetics Estimator

ENKIE predicts enzyme kinetic parameter values and their uncertainties in metabolic networks using Bayesian Multilevel Models.

Key Features:

• Bayesian Multilevel Modeling: Uses Bayesian Multilevel Models to predict values and uncertainties of KM and kcat parameters.
• Categorical Predictors: Employs five categorical predictors to achieve prediction performance comparable to deep learning approaches that use sequence and structure information.
• Calibrated Uncertainty Predictions: Provides calibrated uncertainty predictions and interpretable insights into the main sources of uncertainty.
• Prior Construction for Kinetic Models: Simplifies the construction of priors for Bayesian kinetic models of metabolism.

Scientific Applications:

• Metabolic Network Modeling: Relates metabolite and enzyme abundances to metabolic fluxes through reaction kinetics for dynamic and enzyme cost models.
• Kinetic Parameter Gap-Filling: Addresses the limited availability of measured kinetic parameters and unknown reliability of existing values for enzymes.
• Bayesian Kinetic Model Development: Supports construction of priors for Bayesian kinetic models of metabolism.

Methodology:

Bayesian Multilevel Models with five categorical predictors.