 The soluble carrier hormone binding protein, HBP, is an important factor in the growth of humans and other animals. It can also bind selectively and non-covalently to hormones, making it essential to understand its biological functions and molecular mechanisms. Experimental methods are still time-consuming and expensive, so developing computational methods to accurately and efficiently identify HBP is necessary. A machine learning-based approach was used in this study, where the samples were encoded by using the optimal tripeptide composition obtained from the binomial distribution method. This resulted in an overall accuracy of 97.15% in the five-fold cross-validation test. To make the results more accessible to the scientific community, a user-friendly web server called HBPRED 2.0 was developed, which can be accessed online. This article was authored by GeoXin Tan, Sherhow Lee, Zimei Zhang, and others. We are article.tv, links in the description below.