MiBRP: a predictor using Laplacian support vector machine and sequence information to identify microRNA-binding residues

Conference: BIBE 2019 - The Third International Conference on Biological Information and Biomedical Engineering
06/20/2019 - 06/22/2019 at Hangzhou, China

Proceedings: BIBE 2019

Pages: 6Language: englishTyp: PDF

Personal VDE Members are entitled to a 10% discount on this title

Ma, Xin (School of Statistics and mathematics, Nanjing Audit University, Nanjing, China)

The recognition of microRNA (miRNA)-binding protein residues in proteins enhances our understanding of how miRNAs silence their target genes and several relevant biological processes. We propose a method, miBRP, which combines a novel hybrid feature with a semi-surpervised learning method, i.e., Laplacian support vector machine (LapSVM) algorithm to predict miRNA-binding residues in protein sequences. The hybrid feature is put forward for coding instances which incorporates secondary structure, orthogonal binary vector, and a novel feature, which contributes the most to prediction improvement. The miBRP significantly outperforms the previous method on prediction of miRNA-binding residues.