Using increment of diversity to predict mitochondrial proteins of malaria parasite: integrating pseudo-amino acid composition and structural alphabet |
| |
Authors: | Chen Ying-Li Li Qian-Zhong Zhang Li-Qing |
| |
Institution: | (1) Laboratory of Theoretical Biophysics, School of Physical Science and Technology, Inner Mongolia University, Hohhot, China;(2) Department of Computer Science, Virginia Tech, Blacksburg, VA, USA;(3) Program in Genetics, Bioinformatics, and Computational Biology, Virginia Tech, Blacksburg, VA, USA |
| |
Abstract: | Due to the complexity of Plasmodium falciparum (PF) genome, predicting mitochondrial proteins of PF is more difficult than other species. In this study, using the n-peptide composition of reduced amino acid alphabet (RAAA) obtained from structural alphabet named Protein Blocks as feature
parameter, the increment of diversity (ID) is firstly developed to predict mitochondrial proteins. By choosing the 1-peptide
compositions on the N-terminal regions with 20 residues as the only input vector, the prediction performance achieves 86.86%
accuracy with 0.69 Mathew’s correlation coefficient (MCC) by the jackknife test. Moreover, by combining with the hydropathy
distribution along protein sequence and several reduced amino acid alphabets, we achieved maximum MCC 0.82 with accuracy 92%
in the jackknife test by using the developed ID model. When evaluating on an independent dataset our method performs better
than existing methods. The results indicate that the ID is a simple and efficient prediction method for mitochondrial proteins
of malaria parasite. |
| |
Keywords: | |
本文献已被 PubMed SpringerLink 等数据库收录! |
|