Integration of Data Mining and Meta-analysis of Transcriptomic Data to Identify the Key Genes in Mastitis Resistance in Dairy Cows

Meta-analysis of expression studies is being increasingly employed to identify robust molecular signatures and to gain insights into underlying biological processes. Herein this statistical approach was performed to identify genes significantly involved in response to mastitis in the mammary gland tissue of dairy cows; then the most important genes were identified by data mining approaches. As sampling techniques (in vivo and in vitro) and sampling time after infection were different between experiments, each time was considered as separate study and combined by one tailed p-value meta-analysis method to identify up and down regulated meta-genes. Ten different attribute weighting algorithms were applied to the gene dataset to identify the most important ones. From 892 genes that were significantly differentially expressed in response to mastitis (one tail q< 0/005), three genes (ZC3H12A, PDE4B, CASP4)- functional with the immune system – were selected as the most important genes by attribute weighting algorithms. It should be noted that these genes have not been previously associated with the mastitis diseases in dairy cattle’s.