Related Experiment Video
Updated: May 24, 2026
Elucidating β-1,3-Glucanase and Peroxidase Physicochemical Properties of Wheat Cell Wall Defense Mechanism Against Diuraphis noxia Infestation
Published on: July 26, 2024
Predicting Km values of beta-glucosidases using cellobiose as substrate.
Shao-Min Yan1, De-Qiang Shi, Hao Nong
1State Key Laboratory of Non-food Biomass Enzyme Technology, National Engineering Research Center for Non-food Biorefinery, Guangxi Key Laboratory of Biorefinery, Guangxi Academy of Sciences, Nanning, 530007, Guangxi, China.
Predicting enzyme kinetics is crucial for understanding enzymatic reactions. This study develops models to predict the Michaelis-Menten constant (Km) for beta-glucosidase using its amino acid sequence, aiding enzyme research.
Area of Science:
- Biochemistry
- Enzyme kinetics
- Computational biology
Background:
- The Michaelis-Menten constant (Km) is vital for characterizing enzyme-substrate interactions.
- Experimentally determined Km values are often scarce in scientific literature, hindering enzyme research.
- The increasing number of novel enzymes necessitates predictive methods for kinetic parameters.
Purpose of the Study:
- To develop computational models for predicting the Km of beta-glucosidase with cellobiose as a substrate.
- To utilize the enzyme's primary amino acid sequence information for Km prediction.
- To address the shortage of readily available kinetic data for enzymes.
Main Methods:
- Development of predictive models using the primary structure of beta-glucosidase.
- Utilizing amino acid distribution probability as input features for the models.
- Employing a 20-1 feedforward backpropagation neural network architecture.
Main Results:
- The 20-1 feedforward backpropagation neural network demonstrated the highest accuracy in predicting Km values.
- Amino acid distribution probability proved to be an effective predictor for Km.
- The developed models successfully predicted Km for beta-glucosidase with cellobiose.
Conclusions:
- Predictive modeling based on primary enzyme structure is a viable approach to estimate Km values.
- Neural networks, specifically the 20-1 feedforward backpropagation model, are effective for predicting enzyme kinetic parameters.
- This study provides a valuable tool for researchers needing Km data for beta-glucosidase and similar enzymes.
