This principle has been used to identify molecular pathways Talazoparib PARP inhibitor associated with the transformation of melanocytes into melanomas, and contributes to in silico models of gene-to-gene relationships known as gene networks. In a gene network, a connection between two RNAs implies either co-expression of the two RNAs or the regulation of the abundance of one RNA by the abundance of the other, either directly or via intervening signalling molecules and transcription factors. In gene networks RNAs are usually referred to as ����nodes����, connections between them referred to as ����edges���� and groups of RNAs that are highly correlated with one other are referred to as ����clusters����. There are several types of gene networks that model RNA-to-RNA relationships using different assumptions, ranging from simple non-directional correlation-based methods, sometimes referred to as relevance networks, to complex Bayesian gene networks, which can model directional and synergistic relationships between molecules. Until recently, due to computational limitations, most directional gene network methods could only model interactions between a few hundred genes at a time. However, in 2010, a method to identify whole-genome-scale Bayesian gene networks using massively parallel supercomputers was developed, which is used in this study. In this study we find that the association of tumour clinical features with either individual RNAs or inferred molecular pathway activity is not consistent across published melanoma microarray datasets. Given this lack of consistency, and the consequent difficulty of using data from the diverse melanomas of patients to understand melanoma molecular pathways, we instead take an in vitro functional genomic approach. We generate microarray data from the melanoma cell line A375 exposed to a set of targeted siRNA disruptions, and used these data to identify co-expressed clusters of genes that are strongly conserved between siRNA-treated A375 cells and melanomas from patients. Several of these individual clusters encode proteins with shared cellular functions; we show that those clusters related predominantly to cellular proliferation are significantly associated with the prognosis of metastatic melanoma patients. In several individual published microarray studies of melanomas from patients, sets of genes appear to be differentially expressed in association with three aspects of tumour BAY-60-7550 side effects biology: progression, metastasis and prognosis. We wished to assess whether the genes associated with these clinical features were consistent across the multiple published studies. Therefore, the raw data from several well-designed microarray studies that addressed progression, metastasis and prognosis were retrieved. Quality control assessment indicated that all data was of acceptable quality and re-analysis of each dataset from Table 1 identified sets of differentially expressed RNAs similar to those previously published, although the different studies appeared to vary widely in their statistical power.