The Choice of the Filtering Method in Microarrays Affects the Inference Regarding Dosage Compensation of the Active X-Chromosome

et al. (2011) The Choice of the Filtering Method in Microarrays Affects the Inference Regarding Dosage Compensation of the Active X-Chromosome. PLoS ONE 6(9): e23956. doi:10.1371/journal.pone.0023956 The Choice of the Filtering Method in Microarrays Affects the Inference Regarding Dosage Compensation of the Active X-Chromosome Raphae le Castagne 0 1 Maxime Rotival 0 1 Tanja Zeller 0 1 Philipp S. Wild 0 1 Vinh Truong 0 1 David-Alexandre Tre goue t 0 1 Thomas Munzel 0 1 Andreas Ziegler 0 1 Fran cois Cambien 0 1 Stefan Blankenberg 0 1 Laurence 0 1 Tiret 0 1 Brian P. Chadwick, Florida State University, United States of America 0 Current address: Department of General and Interventional Cardiology, University Heart Center, University Medical Center Hamburg-Eppendorf , Hamburg , Germany 1 1 INSERM UMRS 937, Pierre and Marie Curie University (UPMC, Paris 6) and Medical School , Paris , France , 2 II. Medizinische Klinik und Poliklinik, Johannes-Gutenberg Universita t Mainz, Universita tsmedizin , Mainz, Germany , 3 Institut f u r Medizinische Biometrie und Statistik, Universita t zu Lu beck, Universita tsklinikum Schleswig-Holstein , Lu beck , Germany Background: The hypothesis of dosage compensation of genes of the X chromosome, supported by previous microarray studies, was recently challenged by RNA-sequencing data. It was suggested that microarray studies were biased toward an over-estimation of X-linked expression levels as a consequence of the filtering of genes below the detection threshold of microarrays. Methodology/Principal Findings: To investigate this hypothesis, we used microarray expression data from circulating monocytes in 1,467 individuals. In total, 25,349 and 1,156 probes were unambiguously assigned to autosomes and the X chromosome, respectively. Globally, there was a clear shift of X-linked expressions toward lower levels than autosomes. We compared the ratio of expression levels of X-linked to autosomal transcripts (X:AA) using two different filtering methods: 1. gene expressions were filtered out using a detection threshold irrespective of gene chromosomal location (the standard method in microarrays); 2. equal proportions of genes were filtered out separately on the X and on autosomes. For a wide range of filtering proportions, the X:AA ratio estimated with the first method was not significantly different from 1, the value expected if dosage compensation was achieved, whereas it was significantly lower than 1 with the second method, leading to the rejection of the hypothesis of dosage compensation. We further showed in simulated data that the choice of the most appropriate method was dependent on biological assumptions regarding the proportion of actively expressed genes on the X chromosome comparative to the autosomes and the extent of dosage compensation. Conclusion/Significance: This study shows that the method used for filtering out lowly expressed genes in microarrays may have a major impact according to the hypothesis investigated. The hypothesis of dosage compensation of X-linked genes cannot be firmly accepted or rejected using microarray-based data. - Funding: The Gutenberg Heart Study is funded through the government of Rheinland-Pfalz (Stiftung Rheinland Pfalz fu r Innovation, contract AZ 961-386261/ 733), the research programs Wissen schafft Zukunft and Schwerpunkt Vaskulare Pravention of the Johannes Gutenberg-University of Mainz and its contract with Boehringer Ingelheim and PHILIPS Medical Systems including an unrestricted grant for the Gutenberg Heart Study. The present study was supported by the National Genome Network NGFNplus (contract A3 01GS0833 and 01GS0831) and by a joint funding from the Federal Ministry of Education and Research, Germany (contract BMBF 01KU0908A) and from the Agence Nationale de la Recherche, France (contract ANR 09 GENO 106 01) for the project CARDomics. MR is supported by a grant from the Fondation pour la Recherche Medicale (FDT20101220928). For this particular research paper, Boehringer Ingelheim provided payment of two employees for expression microarray analyses and array purchase. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. . These authors contributed equally to this work. It is widely admitted that in mammals, X-linked genes are upregulated to ensure balanced expression between the X chromosome, present in a single active copy per cell, and autosomes, present in two copies [1]. The hypothesis of dosage compensation first proposed by Ohno in 1967 [2] was supported by recent microarray studies showing that X-linked genes were expressed at similar levels to autosomal genes in mice and humans [35]. However, the molecular mechanism responsible for this compensation is still not understood [6]. The absence of mechanistic interpretation, coupled to the lack of dosage compensation in other taxa [7,8], have spurred speculation about this phenomenon. Recently, Ohnos hypothesis was challenged by a study using RNA sequencing (RNA-Seq) data showing that the ratio of the median expression level of X-linked genes to that of autosomal genes (X:AA) was significantly lower than 1 in different human and mouse tissues [9]. The authors attributed the difference between their findings and previous ones to the fact that RNA-Seq is much more sensitive than microarray to detect small expression differences [911] and that microarray studies are likely to be biased towards an over-estimation of X-linked expression levels as a consequence of the filtering of genes considered to be under the detection threshold of microarray [9]. This controversial finding led us to question the method conventionally used for the analysis of microarray-based expression data. Using data from a large-scale expression study in human monocytes [12], we showed that according to the method used for filtering out the genes prior to analysis, the inference regarding dosage compensation was in the opposite direction. A simulation study further demonstrated that the choice of the most appropriate filtering method was dependent on biological assumptions regarding the proportion of actively expressed genes on the X chromosome and on autosomes and the extent of dosage compensation. Although the limited sensitivity of microarrays does not allow one to go further in resolving this issue, the potential methodological bias arising from using a signalthreshold cutoff in microarray experiments should be kept in mind when comparing expression across loci. Filtering transcripts considered as undetected by microarrays may discard genes that show biologically relevant associations supporting cellular expression In microarray studies, genes whose expression is not significantly different from the background signal are conventionally filtered out prior to analysis. These genes are often inappropriatel (...truncated)