Biostatistical Considerations oftheUseofGenomicDNA Reference inMicroarrays

Usinggenomic DNA as common reference inmicroarray experiments has recently beentestedby different laboratories (2,3,5,7,9,20,24-26). While some reported thatexperimental results ofmicroarrays usinggenomic DNA reference conformed nicely tothoseobtained by cDNA:cDNA co-hybridization method, others acquired poor results. Wehypothesized that these conflicting reports couldberesolved bybiostatistical analyses. Totestit, microarray experimentswere performed in a 4 proteobacterium Shewanellaoneidensis. Pair-wise comparison ofthree experimental conditions was obtained either bydirect cDNA:cDNA co-hybridization, or by indirect calculation through a Shewanella genomic DNA reference. Several majorbiostatistical techniques were exploited toreduce theamountofinconsistency between bothmethods andtheresults were assessed Wediscovered thatimposing theconstraint ofminimalnumberof replicates, logarithmic transformation andrandomerror analyses significantly improved thedataquality. These findings couldpotentially serve as guidelines for microarray dataanalysis using genomic DNA asreference.