Bioinformatics Advance Access originally published online on June 4, 2004
Bioinformatics 2004 20(17):2943-2953; doi:10.1093/bioinformatics/bth338
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Bioinformatics vol. 20 issue 17 © Oxford University Press 2004; all rights reserved.
Bayesian integrated functional analysis of microarray data
1 Rolf Nevanlinna Institute, Department of Mathematics and Statistics, University of Helsinki, P.O. Box 68, FIN 00014, Helsinki, Finland and 2 Division of Human Biology, Fred Hutchinson Cancer Research Centre, Seattle, WA 98109-1024, USA
Received on December 10, 2003; revised on April 30, 2004; accepted on May 16, 2004
Advance Access Publication June 4, 2004
Motivation: The statistical analysis of microarray data usually proceeds in a sequential manner, with the output of the previous step always serving as the input of the next one. However, the methods currently used in such analyses do not properly account for the fact that the intermediate results may not always be correct, then leading to cumulating error in the inferences drawn based on such steps.
Results: Here we show that, by an application of hierarchical Bayesian methodology, this sequential procedure can be replaced by a single joint analysis, while systematically accounting for the uncertainties in this process. Moreover, we can also integrate relevant functional information available from databases into such an analysis, thereby increasing the reliability of the biological conclusions that are drawn. We illustrate these points by analysing real data and by showing that the genes can be divided into categories of interest, with the defining characteristic depending on the biological question that is considered. We contend that the proposed method has advantages at two levels. First, there are gains in the statistical and biological results from the analysis of this particular dataset. Second, it opens up new possibilities in analysing microarray data in general.
Supplementary information: http://www.rni.helsinki.fi/~mab/
Contact: mab{at}rni.helsinki.fi
* To whom correspondence should be addressed.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Pritchard, B. Mecham, R. Dumpit, I. Coleman, M. Bhattacharjee, Q. Chen, R. A. Sikes, and P. S. Nelson Conserved Gene Expression Programs Integrate Mammalian Prostate Development and Tumorigenesis Cancer Res., March 1, 2009; 69(5): 1739 - 1747. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Wilkinson Bayesian methods in bioinformatics and computational systems biology Brief Bioinform, April 12, 2007; (2007) bbm007v1. [Abstract] [Full Text] [PDF]
|
|