Novel Tree-Based Proximity Search with SMOTE and Compositional Indexing Techniques for Protein Domain Identification

Proteins are generally long chains and consist of multiple domains. Domains are distinct structural units of a protein that can evolve and function independently. The accurate and reliable identification of protein domains is a fundamental stage in protein tertiary structure prediction. It not only gives insight into the way proteins work and therefore enhance medicine and drug development, but also reduces the experimental cost of protein structure determination by allowing researchers to work on a set of smaller and informative unit. In this work, we introduce a novel domain identification approach based on protein primary structure information only. We propose a novel tree-based proximity search model trained by amino acid compositional index and physiochemical properties and enhanced by SMOTE (Synthetic Minority Over-sampling Technique).