An Integrative Informatics Approach to Identify Drug Targets and Therapies for Chronic Allograft Nephropathy After Kidney Transplantation.
L. Li, Benjamin P. Readhead, R. Chen, W. Zhang, Barbara T. Murphy, Jan Dudley · Transplantation · 2014
We hypothesized that meta-analysis of independent chronic allograft nephropathy (CAN) gene expression data from multiple centers will produce a robust signature for improved therapeutic discovery. We developed a computational drug repurposing approach to analyze the meta-analysis signature and identify novel therapeutic options. We integrated 5 independent public available human microarray data sets from GEO with newly generated data, yielding a total of 275 samples across 3 different platforms. We re-annotated all probes using AILUN and employed two meta-analysis methods to provide robust estimates of the significant genes. The first method estimates the effect size for each gene in each data set, then summarized the effect size using a fixed effect inverse-variance model. We combined the study-specific effect sizes for each gene into one meta effect-size. For the second method, we performed the Significance Analysis of Microarray (SAM) for each data set, and then used Fisher's exact test to see the enrichment of the genes based on q-values. We employed a drug repurposing approach to identify compounds predicted to ameliorate molecular states associated with CAN based on the concordance of drug-induced gene expression profiles. We identified 85 genes that were significant (FDR<5%) by meta-analysis methods. Toxicological function were enriched for renal damage (p=1.2E-4), kidney failure (p=2.9E-4), renal necrosis/cell death (p=9.6E-4), increased levels of creatinine (p=0.01), renal inflammation (p=9.6E-3), renal tubule injury (p=0.01), and glomerular injury (p=0.02). In addition to immune cell signaling pathways, Tec kinase signaling pathway was also highly enriched with two over-expressed tyrosine kinases and two transmembrane receptors. One inhibitor targeting these kinases is currently being evaluated in a phase I/II for CKD. Furthermore, Fcγ receptor-mediate phagocytosis in macrophages and monocytes pathway was significantly involved by the molecules phosphorylated by those kinases. Drug repurposing analysis identified 5 top compounds predicted to ameliorate CAN, including compounds already shown to ameliorate renal failure or acute ischemic injury in mice. We used an integrated approach of meta-analysis and computational drug repurposing methods for identifying a list of potential drug targets which can potentially ameliorate renal failure.