Format

Send to

Choose Destination
PLoS One. 2013 Jul 10;8(7):e64477. doi: 10.1371/journal.pone.0064477. Print 2013.

Re-annotation of protein-coding genes in the genome of saccharomyces cerevisiae based on support vector machines.

Author information

1
Center of Bioinformatics, School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu, China.

Abstract

The annotation of the well-studied organism, Saccharomyces cerevisiae, has been improving over the past decade while there are unresolved debates over the amount of biologically significant open reading frames (ORFs) in yeast genome. We revisited the total count of protein-coding genes in S. cerevisiae S288c genome using a theoretical approach by combining the Support Vector Machine (SVM) method with six widely used measurements of sequence statistical features. The accuracy of our method is over 99.5% in 10-fold cross-validation. Based on the annotation data in Saccharomyces Genome Database (SGD), we studied the coding capacity of all 1744 ORFs which lack experimental results and suggested that the overall number of chromosomal ORFs encoding proteins in yeast should be 6091 by removing 488 spurious ORFs. The importance of the present work lies in at least two aspects. First, cross-validation and retrospective examination showed the fidelity of our method in recognizing ORFs that likely encode proteins. Second, we have provided a web service that can be accessed at http://cobi.uestc.edu.cn/services/yeast/, which enables the prediction of protein-coding ORFs of the genus Saccharomyces with a high accuracy.

PMID:
23874379
PMCID:
PMC3707884
DOI:
10.1371/journal.pone.0064477
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Public Library of Science Icon for PubMed Central
Loading ...
Support Center