Increasing the Reliability of Protein-Protein Interactions and Protein Function Prediction Based on the Experimental Identification Methods

Protein function prediction is one of the most important problems in the field of proteomics since it leads to determining cell functions. Since the recent computational techniques extract protein function from the protein interactions networks, the reliability of protein interactions is very important task to be considered in the prediction process. Usually, the reliability of interactions based on the experimental or identification methods (laboratory experiments for identifying the interactions) did not be considered in protein function/interaction prediction. In this paper, we propose a strategy for determining the reliability of protein interactions. The technique is based on the number of the experimental methods that have been used in identifying the protein-protein interaction as well as the strength of each experimental method. The reliability of interactions and protein function prediction processes have been improved. The proposed method was applied to the yeast proteome and the results revealed great improvement in increasing the certainty and accuracy for protein function prediction.