Prediction of protein-protein interactions using chaos game representation and wavelet transform via the random forest algorithm.
Genet Mol Res
; 14(4): 11791-805, 2015 Oct 02.
Article
em En
| MEDLINE
| ID: mdl-26436504
Studying the network of protein-protein interactions (PPIs) will provide valuable insights into the inner workings of cells. It is vitally important to develop an automated, high-throughput tool that efficiently predicts protein-protein interactions. This study proposes a new model for PPI prediction based on the concept of chaos game representation and the wavelet transform, which means that a considerable amount of sequence-order effects can be incorporated into a set of discrete numbers. The advantage of using chaos game representation and the wavelet transform to formulate the protein sequence is that it can more effectively reflect its overall sequence-order characteristics than the conventional correlation factors. Using such a formulation frame to represent the protein sequences means that the random forest algorithm can be used to conduct the prediction. The results for a large-scale independent test dataset show that the proposed model can achieve an excellent performance with an accuracy value of about 0.86 and a geometry mean value of about 0.85. The model is therefore a useful supplementary tool for PPI predictions. The predictor used in this article is freely available at http://www.jci-bioinfo.cn/PPI.
Texto completo:
1
Coleções:
01-internacional
Base de dados:
MEDLINE
Assunto principal:
Proteínas de Bactérias
/
Modelos Estatísticos
/
Proteínas de Saccharomyces cerevisiae
/
Mapeamento de Interação de Proteínas
/
Análise de Ondaletas
Tipo de estudo:
Clinical_trials
/
Prognostic_studies
/
Risk_factors_studies
Idioma:
En
Revista:
Genet Mol Res
Assunto da revista:
BIOLOGIA MOLECULAR
/
GENETICA
Ano de publicação:
2015
Tipo de documento:
Article
País de afiliação:
China
País de publicação:
Brasil