Study of decision tree in the application of predicting protein-protein interactions.
- Author:
Xiaolong GUO
1
;
Yan JIANG
1
;
Lu QUI
1
Author Information
1. Institute of System Science and Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China.
- Publication Type:Journal Article
- MeSH:
Algorithms;
Decision Trees;
Fungal Proteins;
chemistry;
Models, Theoretical;
Protein Interaction Domains and Motifs;
Protein Interaction Maps
- From:
Journal of Biomedical Engineering
2013;30(5):952-956
- CountryChina
- Language:Chinese
-
Abstract:
Proteins are the final executive actor of cell viability and function. Protein-protein interactions determine the complexity of the organism. Research on the protein interactions can help us understand the function of the protein at the molecular level, learn the cell growth, development, differentiation, apoptosis and understand biological regulation mechanisms and other activities. They are essential for understanding the pathologies of diseases and helpful in the prevention and treatment of diseases, as well as in the development of new drugs. In this paper, we employ the single decision-tree classification model to predict protein-protein interactions in the yeast. The original data came from the existing literature. Using software Clementine, this paper analyzes how these attributes affect the accuracy of the model by adjusting the predicted attributes. The result shows that a single decision tree is a good classification model and it has higher accuracy compared to those in the previous researches.
