Integrated analysis of the heterogeneous microarray data

Sung Gon Yi,Taesung Park

doi:10.1186/1471-2105-12-s5-s3

Abstract

BackgroundAs the magnitude of the experiment increases, it is common to combine various types of microarrays such as paired and non-paired microarrays from different laboratories or hospitals. Thus, it is important to analyze microarray data together to derive a combined conclusion after accounting for heterogeneity among data sets. One of the main objectives of the microarray experiment is to identify differentially expressed genes among the different experimental groups. We propose the linear mixed effect model for the integrated analysis of the heterogeneous microarray data sets.ResultsThe proposed linear mixed effect model was illustrated using the data from 133 microarrays collected at three different hospitals. Though simulation studies, we compared the proposed linear mixed effect model approach with the meta-analysis and the ANOVA model approaches. The linear mixed effect model approach was shown to provide higher powers than the other approaches.ConclusionsThe linear mixed effect model has advantages of allowing for various types of covariance structures over ANOVA model. Further, it can handle easily the correlated microarray data such as paired microarray data and repeated microarray data from the same subject.

Highlights

As the magnitude of the experiment increases, it is common to combine various types of microarrays such as paired and non-paired microarrays from different laboratories or hospitals
As the cost of producing microarrays has become lower costs and the importance of replication in microarray experiments has been demonstrated by many researchers [1], replicated microarrays are commonly used in microarray experiments
We propose the linear mixed effect (LMe) model for the integrated analysis of the heterogeneous microarray data sets

Summary

Results

Analysis of the liver cancer microarray data We applied the integrated analysis using LMe models, two-stage ANOVA model, and meta-analysis to liver cancer data. The number of identified genes by meta-analysis, ANOVA model, five LMe models are 197, 145, 214, 543, 589, 375, and 114, respectively. For the simulated data sets, we perform the analyses using the meta-analysis, the two-stage ANOVA model and five LMe models. We fit this LMe model by assuming that bil has the covariance structure of Types 1 to 4. Powers and FDRs showed very consistent results for all methods, the variances of tumor tissue and control tissue are assumed to be different. It is probably due to the fact that the meta-analysis allows different variances between two data sets, while others do not

Conclusions

Background

Materials and methods

Type 4

Discussion

Conclusion