Evaluation Of Six Batch Adjustment Methods In Expression Microarray Data And Application Of Gene Co-expression Network In Schizophrenia

The expression microarray is a frequently used approach to study gene expression on a genome-wide scale. However, the data produced by the thousands of microarray studies published annually are confounded by "batch effects," the systematic error introduced when samples are processed in multiple batches. Although batch effects can be reduced by careful experimental design, they cannot be eliminated unless the whole study is done in a single batch. We first stressed batch effects exist and confounded with biological factor in varies microarray platforms, then by outlining the experiment procedures, we pointed out the potential sources of batch effects. A number of programs are now available to adjust microarray data for batch effects prior to analysis. We systematically evaluated six of these programs using multiple measures of precision, accuracy and overall performance. ComBat, an Empirical Bayes method, outperformed the other five programs by most metrics. We also showed that it is essential to standardize expression data at the probe level when testing for correlation of expression profiles, due to a sizeable probe effect in microarray data that can inflate the correlation among replicates and unrelated samples.We then utilized the gene expression microarray to explore thepathogenesis of schizophrenia. Differential gene expression in schizophrenia brain have been pursued by multiple studies, and they yielded a long list of interesting candidate genes, but barely findings can survive multiple testing correction in at least one study and replicated in other studies. This is largely due to strong heterogeneity of gene expression in human brain and maybe minor changes in patients. We hypothesized that coordinated gene expression networks or pathways may have stronger and more robust changes in patient brains. In this study, we used the weighted co-expression network analysis to evaluate expression data of five brain gene expression studies.We first filtered data by strict quality control criteria:besides ComBat for batch correction, we also filtered out probe sets containing SNPs, with detection in less than 90% of samples, and without sufficient annotation.We removed samples of non-Caucasians and outliers from clustering method.A random sample is chosen from replicates.Wefound one module contained genes, MT1X, MT1E, MT1F, MT1G, MT1X and MT2A, which belong to metallothioneins(MT) gene family, were consistently significantly correlated with schizophrenia in five datasets with varied sample sources, and different microarray platforms. This robust change indicated their role in schizophreniaetiology or pathology. MT as one gene family enriched cysteine residues to bind heavy metals, such as zinc, copper, cadmium, mercury, involved in reactive oxygen species protection and stress adaption. Meanwhile, oxidative stress has been suggested to contribute to the pathophysiology of schizophrenia, and zinc plays important roles in nerve development, mood control and preventing cell damage from oxidation; its supplement was considered as one schizophrenia treatment three decades ago. Those evidences are all related with MT's function in central neuron system, indicating MT's potential contribution in schizophrenia pathogenesis.We also tried to explore themechanism of MT's expression alternation in schizophrenia from genetics and epigenetics views, by eQTL method and DNA methylation data, separately.