I'm pretty sure prior art doesn't work that way.

Isn't the standard here whether or not these molecules exist in nature? Honest question.

You can't just assert that it exists. You'd have to show that it exists.

That should be trivial, if you had a large sample and money for sequencing. The consensus in the listservs I'm on is that cDNA patentability will fall next because of these and other inconsistencies (cDNA existing in nature, being a non-natural transformation).

That hinges on the definition of non-natural. Is a nuclear reactor non-natural? Are you sure? http://en.wikipedia.org/wiki/Natural_nuclear_fission_reactor

Just because you're using components that exist in nature that may come together and occasionally produce the result you want - when a human hand enters the picture to do it deliberately and get a controlled result, to gain a certain end, that's qualifies it as non-natural.

cDNAs do not "exist in nature" the way you think they do. Viral RTs typically hijack a tRNA for use as a primer, and there is a specific tRNA footprint site that is necessary to initiate RT specifically at the position of interest to the virus. these footprint sites typically don't exist near the gene you care about.