Melanoma is an interesting cancer to study. First one could surmise based upon many studies that a primary caused of its incidence is the UV radiation which the skin is subjected to. Then there appears to be some genetic correlation. However there are families where one sibling was a lifeguard for five years and managed to get to seventy and where the other sibling dies at 24 and was never exposed to the sun.Thus the true causal relationship may still be some what up in the air. 
However the genetic profiles of certain melanomas, after the fact, and their pathways, are starting to be identified. A recent 
NCI study states:
In one approach, the team searched for mutations that occurred in  multiple patients. They found mutations in the gene BRAF, previously implicated in melanoma, and in 9 other genes. Mutations  in a gene called TRRAP occurred at  the same position in 6 separate people with melanoma.
               To test whether TRRAP is an  oncogene—a gene that can prevent normal cell death and cause cancer—the  researchers disrupted TRRAP in melanoma  cells. As expected for an oncogene, disruption of TRRAP caused an increase in cell death.
               In another approach, the researchers looked for genes  with a higher-than-expected  mutation rate. Of the 16 identified genes,  only BRAF had previously been implicated in melanoma. Another, GRIN2A, is one of the most highly mutated genes associated with  melanoma to date. The identification of GRIN2A, along with other data  from the study, implicate the glutamate signaling pathway in melanoma.
Thus the pathways are becoming a key element. As we have discussed before, there is some experimental evidence of a stem cell model for melanoma. Thus one would ask if the genetic markers are stem cell markets are otherwise. It will become critical to be able to look at markers on a cell by cell basis and not on a larger cross section. Also it will be essential to correlate this with the controlling pathways, again on a cell by cell basis. Large scale cell by cell analysis will be a key driver. Perhaps indicators of cell surface proteins may be a reasonable proxy. Just a thought.