Integration of multi-microarray datasets to identify chronic obstructive pulmonary disease-related miRNAs
Currently, the mechanisms underlying chronic obstructive pulmonary disease (COPD) remain unclear. As potential biomarkers, microRNAs (miRNAs), which modulate the levels of specific genes and proteins, are important for enhancing our understanding of the mechanisms behind COPD. Although there have been a number of miRNA expression profiling analyses strategies used to document miRNA expression changes during physiological and pathological processes or used to identify differentially expressed miRNAs in disease or control samples, the study results have been inconsistently replicated using different datasets. For this reason, many findings cannot be well synthesized and interpreted. To address this issue, we used a multiple co-inertia analysis (MCIA) method to extract potential COPD-related miRNAs using three COPD microarray datasets. The results showed that miR-223, miR-132, and miR-199a-5p are obviously associated with COPD, and these results are consistent with the highly significant differentially-expressed miRNAs that were observed across three microarray datasets. Moreover, when miR-223, miR-132, and miR-199a-5p are taken as predictors to classify the samples of three datasets, the pooled sensitivity and specificity is 0.96 and 0.75, respectively, thereby suggesting that these three miRNAs can effectively distinguish COPD patients and controls.
