iHBP-DeepPSSM: Identifying hormone binding proteins using PsePSSM based evolutionary features and deep learning approach

Hormone binding proteins (HBPs) are soluble carrier proteins that can non-covalently and selectively interact with the human hormone. HBPs plays a significant role in human life, but its functions are still unclear. Conventional experimental approaches are deemed ineffective due to its high time processing time and cost. Due to the rapid increases in protein sequences, truly characterization of HBPs has become a challenging task for investigators. Measuring the effectiveness of HBPs on the human body, an accurate and reliable intelligent model is desirable for the identification of hormone-binding proteins. In this paper, high discriminative evolutionary features are extracted using a position-specific scoring matrix (PSSM) and Pseudo position-specific scoring matrix (PsePSSM). On the other hand, to reflect the intrinsic correlation and sequence order information, Series Correlation Pseudo Amino Acid Composition (SC-PseAAC) is also applied. Furthermore, to reduce the computational time and to eradicate irreverent and noisy features, the Sequential forward selection and Support Vector Machine (SFS-SVM) based ensemble approach is applied to select optimal features. Furthermore, various diverse nature of learning algorithms is utilized to select the best operational engine for our proposed model. After evaluating the empirical outcomes, deep neural network using optimal PsePSSM features obtained remarkable results. Whereas, our proposed iHBP-DeepPSSM achieved an accuracy of 94.41%, 92.31%, and 90.48% using the training dataset, and independent datasets (S2 and S3), respectively. It is observed that iHBP-DeepPSSM shows an outstanding improvement compared to literature methods. It is expected that the developed model may be played a useful role in research academia as well as proteomics and drug development. The source code and all datasets are publicly available at https://www.github.com/salman-khan-mrd/iHBP-DeepPSSM