Statistical Test of Expression Pattern (STEPath): a new strategy to integrate gene expression data with genomic information in individual and meta-analysis studies

BMC Bioinformatics

Methodology Article

Paolo Martini¹  Stefano Cagnin¹  Gerolamo Lanfranchi¹  Chiara Romualdi²  Davide Risso³  Gabriele Sales³ 

[1] CRIBI Biotechnology Centre, University of Padova, via U. Bassi 58/B, 35121, Padova, Italy;Department of Biology, University of Padova, via U. Bassi 58/B, 35121, Padova, Italy;Department of Biology, University of Padova, via U. Bassi 58/B, 35121, Padova, Italy;Department of Statistical Science, University of Padova, via C. Battisti 241, 35121, Padova, Italy;

关键词: Acute Lymphoblastic Leukaemia;    Limb Girdle Muscular Dystrophy;    Imbalanced Region;    Skeletal Muscle Disease;    Chip Definition File;   

DOI  :  10.1186/1471-2105-12-92

 received in 2010-10-08, accepted in 2011-04-11,  发布年份 2011

来源: Springer

【 摘 要 】

BackgroundIn the last decades, microarray technology has spread, leading to a dramatic increase of publicly available datasets. The first statistical tools developed were focused on the identification of significant differentially expressed genes. Later, researchers moved toward the systematic integration of gene expression profiles with additional biological information, such as chromosomal location, ontological annotations or sequence features. The analysis of gene expression linked to physical location of genes on chromosomes allows the identification of transcriptionally imbalanced regions, while, Gene Set Analysis focuses on the detection of coordinated changes in transcriptional levels among sets of biologically related genes.In this field, meta-analysis offers the possibility to compare different studies, addressing the same biological question to fully exploit public gene expression datasets.ResultsWe describe STEPath, a method that starts from gene expression profiles and integrates the analysis of imbalanced region as an a priori step before performing gene set analysis. The application of STEPath in individual studies produced gene set scores weighted by chromosomal activation. As a final step, we propose a way to compare these scores across different studies (meta-analysis) on related biological issues. One complication with meta-analysis is batch effects, which occur because molecular measurements are affected by laboratory conditions, reagent lots and personnel differences. Major problems occur when batch effects are correlated with an outcome of interest and lead to incorrect conclusions. We evaluated the power of combining chromosome mapping and gene set enrichment analysis, performing the analysis on a dataset of leukaemia (example of individual study) and on a dataset of skeletal muscle diseases (meta-analysis approach).In leukaemia, we identified the Hox gene set, a gene set closely related to the pathology that other algorithms of gene set analysis do not identify, while the meta-analysis approach on muscular disease discriminates between related pathologies and correlates similar ones from different studies.ConclusionsSTEPath is a new method that integrates gene expression profiles, genomic co-expressed regions and the information about the biological function of genes. The usage of the STEPath-computed gene set scores overcomes batch effects in the meta-analysis approaches allowing the direct comparison of different pathologies and different studies on a gene set activation level.

【 授权许可】

Unknown   
© Martini et al; licensee BioMed Central Ltd. 2011. This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.