Identification of Hormone Binding Proteins via PSeACC

Abstract

Hormone binding protein (HBP) is analogous to a soluble protein carrier and can interact with human hormones non-covalently and selectively. HBP plays an imperative function in the growth of life, but its role remains uncertain. The first step in advancing the study of their work and recognizing their biological process is the correct identification of HBPs. It is difficult, however, to correctly classify HBPs from via conventional biochemical experiments, due to high experimental costs and long experimental time, more and more proteins. Meanwhile, experimental methods are still labor-intensive and cost-effective to identify HBP, developing computational methods to identify HBP accurately and efficiently is crucial. In this analysis, a method based on machine learning was suggested to classify the HBP during which the samples were encoded using the optimal composition of tripeptides obtained by supporting the binomial distribution method. The suggested approach yielded an overall precision of 97.15 percent in the 5-fold cross-validation test. A new technique for recognizing HBPs is provided by this report.