iGHBP: Growth hormone binding proteins #

Introduction #

Growth hormones (GH), and its concentration are the important factor for various diseases control mechanism. Particularly the GH synthesis and GH-interaction with the signaling molecule and understanding the mechanism are the most important for treating the diseases conditions. Since the experimental screening methods are highly expensive to examine the putative candidates, so to select the near-accurate/ similar candidate the computational methods are raised upon the existing evident datasets. In the continuation of those methods, iGHBP is recently published method, which outperformed other existing method i.e., HBPred.

Machines #

1) Random forest 
2) Extremely randomized tree
3) Support vector machine 
4) Adaptive boosting
5) Gradient boosting

Features #

1) Amino Acid Composition (AAC)
2) Di-peptide composition (DPC)
3) Composition-Transition-Distribution (CTD)
4) Amino-acid indices (AAI)
5) Physio-chemical properties (PCP)

Data sets #

1) Benchmark dataset
    a.  123 GHBPs and 123 non-HBPs  
2) Validation dataset 
    a.  31 GHBPs and 31 non-HBPs

Results #

- In this method development they used five machine, based on author assessment, the ERT machine performed well
- The combinations of DPC and AAI, supported to obtain the best predictions

Web-application #

• http://www.thegleelab.org/iGHBP/

Reference #

1. Basith, S., Manavalan, B., Shin, T.H., and Lee, G. (2018).
2. iGHBP: Computational identification of growth hormone binding proteins from sequences using extremely randomised tree. Computational and Structural Biotechnology Journal 16, 412-420.
3. doi: https://doi.org/10.1016/j.csbj.2018.10.007.