HETEROGENEITY OF GLOBAL GENE EXPRESSION MICROARRAY DESIGNS IN DETECTING DIFFERENTIALLY EXPRESSED GENES

DOUGBA DAGO NOEL¹*, FERRARINI ALBERTO², XUMERLE LUCIANO³, MORI ANTONIO⁴, DELLEDONNE MASSIMO⁵, MALERBA GIOVANNI^6
1Department of Biotechnology, University of Verona, Italy Strada le Grazie 15, Ca vignal 1, 37134, Verona Italy
²Department of Biotechnology, University of Verona, Italy Strada le Grazie 15, Ca vignal 1, 37134, Verona Italy
³Department of Biotechnology, University of Verona, Italy Strada le Grazie 15, Ca vignal 1, 37134, Verona Italy
⁴Department of Neurological, Biomedical and Movement Sciences University of Verona, Strada Le Grazie 8, 37134, Verona, Italy
⁵Department of Biotechnology, University of Verona, Italy Strada le Grazie 15, Ca vignal 1, 37134, Verona Italy
⁶Department of Neurological, Biomedical and Movement Sciences University of Verona, Strada Le Grazie 8, 37134, Verona, Italy
* Corresponding Author : dgnoel7@gmail.com

Received : 25-05-2016     Accepted : 01-08-2016     Published : 21-08-2016
Volume : 7     Issue : 2       Pages : 349 - 357
Int J Bioinformatics Res 7.2 (2016):349-357

Keywords : Microarray designs, RNA-Seq, Differentially Expressed (DE) gene, Oligonucleotide Probes, Vitis vinifera
Conflict of Interest : None declared
Acknowledgements/Funding : The authors thank Genomic Center of the University of Verona (Italy) for providing microarrays and RNA-Seq gene expression row data as well as for their technical support for the present work.
Author Contribution : None declared

- Microarray is widely used for gene expression studies by many laboratories worldwide. Microarrays vary for the type and number of oligonucleotide probes implemented and for the procedure to subtract background noise (BS) and normalize data (DN) among samples consenting to make these reliable tools somewhat heterogeneous, as heterogeneity may play an important role identifying differentially expressed (DE) genes in global gene expression studies. We essayed four different microarray design strategies based on either single replicate or multiple probes per gene model transcript and on different probes size (long and/or short) to analyze two Vitis vinifera berry developmental stages. Microarray data were processed basing on 20 different BS-DN arrangements. In addition, Vitis vinifera RNA samples were also analyzed by sequencing-based methods generally referred to as RNA-Seq whose results were used as reference values. Microarray performances in detecting DE genes were evaluated by several measures comprising correlation between estimated fold-change values, classification functions and the Area Under Curve (AUC) of receiver operating characteristic (ROC) curves. The number of DE genes changed from one microarray design to another, suggesting their heterogeneous performances in gene expression differential analysis. However our findings suggested a good agreement between microarrays and RNA-Seq technologies for gene expression level higher than 10 fpkm discriminating differentially (DE) expressed genes. The present results warn researchers that even if different microarray designs can lead different results, both RNA-Seq and array approaches can exhibit comparable performance in gene expression analysis for higher expressed gene. Then, the present survey provided a powerful methodology helping researchers choosing microarrays and/or RNA-Seq approaches in their transcriptomic studies.