I'd be surprised if the appellate court could give you a set of criteria for what exactly makes "a distinctive chemical identity and nature".

Inside the body, your DNA is packed up into chromatin by wrapping around other proteins called histones, and then coiling those thicker strands around itself a few more times. One could make the argument that it's shape is different from isolated DNA, which doesn't have the histones, and is going to be much closer to a short string. However, trying to make the claim that this is a chemically distinct molecule is dubious at best, since the chemical structure and identity of the DNA molecule is unchanged.

There are, however, chemical modifications to the DNA itself that that may not get preserved in the extraction technique (but that really depends on your technique), and almost certainly won't be preserved if you need to amplify your sample to sequence it. The most common way is to put a methyl group (CH3) onto a particular part of the DNA molecule. Single-molecule sequencing techniques exist (but aren't comparably cheap to Illumina or SOLID sequencing yet) that would preserve this chemical structure.

Finally, Illumina sequencing (and probably SOLID too, though I'm not sure) involves adding on extra sequences at the end of your DNA fragments, which allow the sequencer to start it's work. I'd doubt if the patent covers this approach, so doing high-throughput genome sequencing like this ought to be fine, based on their "markedly different structure" standard, but who knows what the court would actually say.

But the appellate decision Friday rejected Judge Sweet’s reasoning, saying that since DNA is a chemical, the chemical structure is what matters and that “informational content is irrelevant to that fact.”

This paragraph is the most troubling to me. In the most obvious definition of the terms, the informational content and the chemical structure of DNA are two different ways of saying the same thing. It would be like saying, "the shape of the splotches of ink on this paper are what's being patented, and the informational content is irrelevant"

Isolated DNA is artificially synthesized DNA derived from a biologically produced mRNA transcript.

If you pretend the mRNA is object code, DNA is source code, and isolated DNA is reverse-engineered object code, you have a pretty solid metaphor. Oh, and comments = introns.

You're using "Isolated DNA" to mean cDNA (a point that I think needs clarifying; you can isolate DNA from cells through other methods). Your metaphor doesn't seem particularly solid to me; it would be more accurate to say that DNA is source code, proteins (and hence enzymes) are object code, and mRNA is code that has been pulled out from disk (the genome), had its comments (introns, as you point out) removed, but is not yet compiled. The computing metaphor obviously breaks down; the best way to characterize cDNA I can think of is as code that's been intercepted on its way to the compiler and placed back into storage - for it to be reverse-engineered object code, it would need to be derived from proteins, which it is not.

We could always call the proteins the binaries, that is to say the output of the linker!

Either way, I was just trying to phrase it in terms familiar to programmers. No metaphor is perfect.

That metaphor really only holds up if your "compiler" does nothing but remove comments (and introduce errors at a low rate).

Obviously DNA is low level. This means it must be assembler- and assembly often is as simple as removing comments and translating "written" code into "machine" code.

Yes, if your "compiler" translates "assembly with comments" to "assembly without comments," then your metaphor is apt.

Update: After reconsidering, I take it back. Introns are important, especially intron boundaries. So, my new assertion: so long as your "compiler" translates "assembly with comments" into "assembly with the same comments intact," then I agree that your metaphor is apt.

So... compiler directives? ;)