Journal of Engineering and Applied Sciences
ISSN: Online 1818-7803ISSN: Print 1816-949x
Abstract
Machine learning-based methods are the most prominently employed in methods in the development of novel protein fold recognition tools. The most recent fold recognition method was developed by combining the four descriptors (e-Values) of Position Specific Iteration BLAST (PSI BLAST), reverse PSI-BLAST (RPS-BLAST), alignment of Secondary Structure Elements (SSE) and PROSITE motifs. In this present study, we emphasized to improve the fold recognition methods by including gene-ontology terms as additional descriptors which can aid in the determination of function based predictions. This method of descriptor combinations have resulted high sensitivity in determining the protein folds when compared to the methods developed with single descriptors. Also, the inclusion of GO-term descriptor have highly increased the sensitivity of the methods in fold recognition which significantly envisages the usage of GO-terms as prominent descriptors that can be employed in the protein fold predictions.
How to cite this article:
E. Loganathan, K., Dinakaran, S. Gnanendra and P. Valarmathie. Function Based Predictions of Protein Fold Recognition using Go-Term.
DOI: https://doi.org/10.36478/jeasci.2017.7534.7538
URL: https://www.makhillpublications.co/view-article/1816-949x/jeasci.2017.7534.7538