Bayesian Pathway Analysis (BPA) Web Server
Bayesian Pathway Analysis (BPA) is a computational method to identify significant pathways from microarray data with samples from test and control classes.
In this method, Bayesian Network (BN) models are constructed by using existing pathways from pathway information databases. A score, which is a measure of probability of data given the fixed network structure, is calculated using Bayesian Scoring Criterion or other scoring metrics. Confidence in this score (p-value) is measured by bootstrapping test. p-value indicates the significance of the pathway, which, therefore, means that an alteration in the state of the pathway under consideration is detected between test case and control case. FDR values are calculated to do correction for multiple hypotheses testing.
BPA Publications
- Korucuoglu, Melike, Senol Isci, Arzucan Ozgur, and Hasan H. Otu. "Bayesian pathway analysis of cancer microarray data." (2014): e102803.
- Isci, Senol, Cengizhan Ozturk, Jon Jones, and Hasan H. Otu. "Pathway analysis of high-throughput biological data within a Bayesian network framework." Bioinformatics 27, no. 12 (2011): 1667-1674.
For questions please contact Hasan H. Otu