A-MADMAN: Annotation-based microarray data meta-analysis tool

Andrea Bisognin,Davide Risso,Francesco Ferrari,Stefania Bortoluzzi,Silvio Bicciato,Chiara Romualdi,Alessandro Coppe

doi:10.1186/1471-2105-10-201

Abstract

BackgroundPublicly available datasets of microarray gene expression signals represent an unprecedented opportunity for extracting genomic relevant information and validating biological hypotheses. However, the exploitation of this exceptionally rich mine of information is still hampered by the lack of appropriate computational tools, able to overcome the critical issues raised by meta-analysis.ResultsThis work presents A-MADMAN, an open source web application which allows the retrieval, annotation, organization and meta-analysis of gene expression datasets obtained from Gene Expression Omnibus. A-MADMAN addresses and resolves several open issues in the meta-analysis of gene expression data.ConclusionA-MADMAN allows i) the batch retrieval from Gene Expression Omnibus and the local organization of raw data files and of any related meta-information, ii) the re-annotation of samples to fix incomplete, or otherwise inadequate, metadata and to create user-defined batches of data, iii) the integrative analysis of data obtained from different Affymetrix platforms through custom chip definition files and meta-normalization. Software and documentation are available on-line at .

Highlights

Available datasets of microarray gene expression signals represent an unprecedented opportunity for extracting genomic relevant information and validating biological hypotheses
A-MADMAN approach The purpose of this paper is to present A-MADMAN, an open source web application for the meta-analysis of Affymetrix data contained in Gene Expression Omnibus (GEO)
A-MADMAN addressees several of previously stated open issues in the meta-analysis of gene expression data allowing the integrative analysis of data obtained from different Affymetrix platforms through custom chip definition files and meta-normalization and the sharing of analysis flows and results

Summary

Introduction

Available datasets of microarray gene expression signals represent an unprecedented opportunity for extracting genomic relevant information and validating biological hypotheses. Major repositories of microarray data, e.g. Gene Expression Omnibus [1], ArrayExpress [2], or Stanford Microarray Database [3], are exceptionally rich mines of genomic information and exploiting their content, through metaanalysis, represents an unprecedented opportunity to improve the interpretation and validation of expression studies. Meta-analysis of large microarray expression datasets allows researchers to confirm biological hypotheses, formulated from results of a study, in a relatively inexpensive way, i.e. using data independently obtained in another laboratory, without the need of novel experiments. Challenges of integrative analysis of expression data In recent years, different strategies to combine results from independent but related studies have been proposed. Data combination encompasses the direct comparison of different studies, is applicable only when expression profiles have been obtained using the same array technology (e.g. Affymetrix, Agilent, Illumina, etc.) and requires an ad-hoc normalization step of the raw data files

Objectives

Results

Discussion

Conclusion