ABOUT   |   AIMS and SCOPE   |   AUTHOR INSTRUCTIONS   |   EDITORIAL BOARD   |   ARCHIVE   |   CURRENT ISSUE   |   SUBMIT MANUSCRIPT ONLINE  |
bioinfo issue

This Article

Copyright © 2015 | AIZEON publishers | All rights reserved

 Open Access
 Abstract
 Full Text PDF Free

 

Services

Articles by corresponding author in PubMed

Articles by corresponding author in Google Scholar

 

 

aizeon

 



 

International Journal of Computational Bioinformatics and In Silico Modeling
2015: Volume-4 Issue-3
ISSN: 2320-0634

>>Back to Issues

ABSTRACT   REFERENCES  
International Journal of Computational Bioinformatics and In Silico Modeling 4(3) 2015: 667-672

Protein Function Prediction Using Hypergeometric Distance with Background Frequency from Protein-Protein Interaction Network



Md. Khaled Ben Islam1*, Julia Rahman2, Md. Al Mehedi Hasan2 and Md. Abdur Rahim1

1 Department of Computer Science & Engineering, Pabna University of Science & Technology, Pabna, Bangladesh
2 Department of Computer Science & Engineering, Rajshahi University of Engineering & Technology, Rajshahi, Bangladesh

* Corresponding Author

ABSTRACT

Due to the availability of large scale experimental data of protein-protein interaction, it remains a challenging task to predict the protein’s function from those high-throughput datasets. Widespread presence of false positive interaction and annotation error in public protein databases are the primary barrier in this case. Considering this fact, we work on network-based statistical approach to predict protein function. Our work is based on the insight that if two proteins have significantly larger numbers of common interaction partners in the experimental network than what is expected from a random network, they have close functional associations and most frequently occurred function in the network has higher probability to be found in the neighbourhood of a protein. First we pre-process both BioGRID and DIP interaction datasets, and then employ the chi-square scoring method with and without considering hypergeometric distance of each interacting pairs. We performed leave-one-out validation to evaluate prediction performance in both cases and find that considering hypergeometric distance improves the performance significantly.

 


Copyright © 2015 | AIZEON publishers | All rights reserved

..........................................................................................................................................................................................................

Citation: Md. Khaled Ben Islam et al. (2015). Protein Function Prediction Using Hypergeometric Distance with Background Frequency from Protein-Protein Interaction Network. Int J Comput Bioinfo In Silico Model 4(3): 667-672

..........................................................................................................................................................................................................