Enzyme sub-functional class prediction based on sequence-structure knowledge

Abstract

Enzyme sequences can be classified based on their structure similarity and/or common evolutionary origin called the sub-functional classes. Information on enzyme sub-functional class is readily available, easing the protein structure and enzyme function probing. ENZYME data-base and UniProt/Swiss-Prot are two prominent classification schemes used to assign the sub-functional class of enzymes. Both schemes determine the subclasses manually based on known main functional classes of enzyme. However, the quantity of known protein sequences is growing exponentially with respect to the quantity of known enzyme functional class. As pointed in the previous literature, it is estimated that only 3-4% of known protein sequences can be assigned to corresponding known sub-functional classes. The fact that this is a tedious and time consuming manually-determined method has further limited the enzyme sub-functional class assignment. In hybrid methods, the combination of sequence-structure knowledge in enzyme sub-functional class prediction allows for proper identification of true positives and true negatives for each query sequence. Besides, with the growing number of unannotated sequences, association of a new sequence to an enzyme of known structure can be a significant step towards the identification of its biological role in enzymatic function.