The pinky bone found at the cave at Denisova in Siberia produced both nuclear DNA and mitochondrial DNA, allowing scientists to sequence (imperfectly) much of the Denisovan genome.
Will this now happen for all ancient fossils?
The simple answer is no, and here is why. I will explain the issues first, and then briefly tell you why they indicate that extracting nuclear DNA from a fossil is very unlikely.
Nuclear DNA is the DNA that varies generation by generation. We also have tiny power plants in each of our cells called mitochondria. Mitochondria have their own DNA, which does not change from generation to generation except by chance mutation. All humans get their mitochondrial DNA from their mother only.
Mitochondrial DNA is more easily preserved than nuclear DNA.
It was a stroke of fortune that the Denisova pinky provided good nuclear DNA, having partially to do with the frozen tundra and partially to do with unknown causes.
No matter the reason, Krause et al describe the unusual condition of the fossil in this way:
In other words, the DNA in the specimen is unusually good in that 70% of the DNA obtained is "endogenous," or DNA that was inherited purely from ancestors. ("Endogenous" means "growing from the inside.") Microbial DNA would be DNA that invaded the Denisovan genome from viruses and other microbes, which is more likely to be unable to be dated and common to many species.
Further, when a scientist is trying to compare DNA from a specimen such as Neandertal or Denisova to the human genome, it's nice to have long pieces of DNA. You can probably imagine that if you had only two or three base pairs, it would match millions of places on our DNA. The larger the length of DNA, the less places that section matches. You really want strands that match only one section of our DNA. The more of those you have, the more likely you are to be able to reconstruct the DNA.
As a result, despite having enough DNA sections to reconstruct the entire genome twice over, Neanderthal DNA is only about 80% certain because there is too much question about some sections represented by only smaller pieces of DNA.
As Dr. Reich, one of the genome researchers puts it:
Both the Denisovans and Neanderthals are relatively recent fossils that prompt great interest as ancestors or relatives of modern humans. Neither goes back further than half a million years, and most Neanderthal remains are much more recent than that. (And, of course, the only Denisovan fossils are from 30 to 50 thousand years ago.)
We were lucky with the Denisova fossils. Neanderthals are represented by hundreds of finds, some of them full skeletons. There are many opportunities to extract Neanderthal remains.
Dinosaurs, on the other hand, are from 225 to 65 million years ago, three orders of magnitude (1,000 or 103 times) older than the Denisova and most Neandertal finds. If the best we can do with hundreds of Neandertal specimens from 100,000 years ago and usually much less, then how will we reproduce the genome of individual species of dinosaurs that are a thousand times older represented by far fewer specimens?
Jurassic Park is unlikely to happen, despite the isolation of proteins from a hadrosaur and Tyrannosaurus Rex. Even if they manage to isolate a cell, it is almost certainly impossible that there will ever be enough DNA, sufficiently well-preserved, to reconstruct any decent portion of dinosaur DNA.