Identifying the Enzymatic Mode of Action for Cellulase Enzymes by Means of Docking Calculations and a Machine Learning Algorithm

Docking calculations have been conducted on 36 cellulase enzymes and the results were evaluated by a machine learning algorithm to determine the nature of the enzyme (i.e. endoor exoenzymatic activity). The docking calculations have also been used to identify crucial substrate-enzyme interactions, and establish structure-function relationships. The use of carboxymethyl cellulose as a docking substrate is found to correctly identify the endoor exobehavior of cellulase enzymes with 92% accuracy while cellobiose docking calculations resulted in an 86% predictive accuracy. The binding distributions for cellobiose have been classified into two distinct types; distributions with a single maximum or distributions with a bi-modal structure. It is found that the uni-modal distributions correspond to exotype enzyme while a bi-modal substrate docking distribution corresponds to endotype enzyme. These results indicate that the use of docking calculations and machine learning algorithms are a fast and computationally inexpensive method for predicting if a cellulase enzyme possesses primarily endoor exotype behavior, while also revealing critical enzyme-substrate interactions.