The last decade was hallmarked by the exceeding technological progress in genome research. As a result genetic alterations such as activating mutations in e.g. BRAF or NRAS, and focal amplifications in e.g. c-KIT, MITF or NEDD9 have been identified as drivers of melanoma, enabling the development of novel targeted therapies. The clinical success, however, has been limited so far. Therefore it is indispensable to uncover further genes driving tumorigenicity and metastasis in human melanoma. We identified gene signatures with potential biological and/or clinical relevance in human melanoma using gene expression profiling and array comparative genomic hybridization. We integrated copy number amplification and gene expression data from sample-matched melanoma specimen to identify recurrent non-random copy number amplifications associated with increased transcript level. Y14, residing at 1q21.1, has the highest correlation of gene dose and transcript levels. Gene expression signatures in particularly high Y14 expressing samples are primarily associated with tumorigenicity, survival and metastasis. In vitro we observed Y14s requirement for melanoma cell proliferation, survival and migration. We also identified a molecular predictor of metastasis in patients with primary cutaneous melanoma. We used gene expression data from primary melanomas to generate a class comparison dataset between the non-metastatic and metastatic condition. For class prediction of metastasis during follow-up we established a second independent dataset from primary cutaneous melanomas. Based on a multivariate Cox regression analysis followed by leave-one-out cross validation we identified EVL and CD24 as gene set giving the best predictive value for metastasis development. Additionally, the gene set is a potential independent predictor of metastasis in primary cutaneous melanomas. New therapeutic strategies for example combination therapy might be a useful approach to combat this devastating disease. By gene set enrichment analysis, we found an enrichment for the expression of genes associated with cell cycle and the differential expression of its member PLK-1 in primary and metastatic human melanomas as compared to melanocytic nevi. Decreased PLK-1 levels led to caspase-3/8 dependent, p53-independent, induction of apoptosis. Decreased PLK-1 expression subsequently to MEK/ERK pathway inhibition is a result of G1 cell cycle arrest, rather than representing a direct regulation of PLK-1. Notably, we determined an additive effect in suppressing the cell viability through combination of MEK/ERK pathway inhibition and interference with PLK-1 expression. Using qRT-PCR we detected SHH-GLI pathway activation in melanoma cell lines. We observed a significantly higher expression of GLI1 in human primary melanoma samples harboring BRAFV600E mutation compared to BRAFWT tissues. We specifically inhibited BRAFV600E in human melanoma cell lines and noticed a decreased expression of GLI1 and pERK1/2. Inhibition of the SHH-GLI pathway suppressed cell proliferation, induced cell cycle arrest, induced apoptosis and reduced melanoma tumor growth in vivo. Altogether, we propose four approaches guiding the identification and/or characterization of gene expression signatures driving melanoma tumorigenicity, survival and metastasis. We suggest a completely novel strategy for therapeutic intervention in a subtype of melanomas showing copy number amplifications at 1q21.1 and increased expression of Y14. Additionally we present a molecular predictor of metastasis in primary cutaneous melanomas, which may also have translational impact upon further evaluation in independent datasets. Finally we suggest two potential therapeutic approaches in human melanoma. On the one hand combination therapy, inhibiting PLK-1 and MEK/ERK signaling and on the other hand specifically targeting the SHH-GLI signaling pathway using the potent and selective SMOOTHEND inhibitor NVP-LDE225.