Efficient Prediction of Progesterone Receptor Interactome Using a Support Vector Machine Model

International Journal of Molecular Sciences

Ji-Long Liu¹  Ying Peng²  Yong-Sheng Fu² 

[1] College of Veterinary Medicine, South China Agricultural University, Guangzhou 510642, China;

关键词: protein–protein interaction;    support vector machine;    progesterone receptor;   

DOI  :  10.3390/ijms16034774

来源: mdpi

【 摘 要 】

Protein-protein interaction (PPI) is essential for almost all cellular processes and identification of PPI is a crucial task for biomedical researchers. So far, most computational studies of PPI are intended for pair-wise prediction. Theoretically, predicting protein partners for a single protein is likely a simpler problem. Given enough data for a particular protein, the results can be more accurate than general PPI predictors. In the present study, we assessed the potential of using the support vector machine (SVM) model with selected features centered on a particular protein for PPI prediction. As a proof-of-concept study, we applied this method to identify the interactome of progesterone receptor (PR), a protein which is essential for coordinating female reproduction in mammals by mediating the actions of ovarian progesterone. We achieved an accuracy of 91.9%, sensitivity of 92.8% and specificity of 91.2%. Our method is generally applicable to any other proteins and therefore may be of help in guiding biomedical experiments.

【 授权许可】

CC BY   
© 2015 by the authors; licensee MDPI, Basel, Switzerland.