By the way, I’ve been meaning to weigh in on the technical side of this. I’ve heard of people doing studies now where they take some pond water and basically sequence all of the DNA in it, discovering new species of bacteria, or at least that there were species there that hadn’t been known to be there before.
I don’t know exactly how these experiments are designed, but if it were me doing it, and I found some previously unknown DNA, here’s how I’d handle it. First of all, a sequence would be flagged as “unknown” if there were no matches for it in all of the gazillions of sequences currently available at various databases. The next step (actually, this would be done at the same time) would be to find all of the closest matches - sequences that are similar, if not identical. This should, hopefully, give you some idea of (a) what kind of sequence you’re looking at - is it a protein-coding gene? A regulatory region? A chunk of repeat DNA? and (b) what sort of organism it came from. If the top ten hits are all from types of duck, odds are you have a chunk of a new species of duck’s DNA there.
So let’s say you actually have some DNA that actually came from Nessie - that is, a truly previously unknown species. I’m not sure exactly how that would show up in a database search. A lot would depend on just how distantly related the organism is to its relatives in the databases, and how many mutations have occurred. As a WAG, I’d suggest that the best hits would be from its nearest relatives, possibly some reptiles and/or birds, if we’re going with Nessie being a dinosaur. If you had enough DNA, you could construct a phylogenetic tree, showing exactly how Nessie is related to other species we know about, but you’d need a pretty substantial chunk of the genome to have any confidence in the result.
