A comparison of human DNA to 12 other animals shows we share more than our genes and helps show that people are more closely related to rats than to cats, scientists reported on Wednesday.

The survey also adds to the argument that so-called 'junk' DNA is nothing of the sort, but must do something important because it stays virtually identical across many species.

It also supports what is becoming increasingly clear -- that the stretches of DNA we call genes are only a small part of the genetic story.

The research team at the National Human Genome Research Institute and several universities compared the same stretch of DNA in a chimpanzee, baboon, cat, dog, cow, pig, rat, mouse, chicken, zebrafish and two species of pufferfish -- the Japanese delicacy Fugu and Tetraodon -- with human DNA.

In people, this stretch of DNA it is a much-studied genetic region that contains the CFTR gene that, when mutated, causes cystic fibrosis.

'It provides some pretty definitive evidence that we are indeed closer to rodents than we are to carnivores,' Dr. Eric Green, scientific director of the NHGRI and leader of the study, said in a telephone interview.

'Our data really puts the nail in the case. In the sequence you can find changes in the genome that clearly occurred in both humans and rodents but did not occur in others.'

The changes come in repetitive sequences of DNA that, until just a few years ago, were believed to be junk -- useless stretches of trash that somehow got saved.

DNA is very difficult to interpret. Its long, long code is built on just four nucleotides -- the compounds known by the abbreviations A,C,T and G.

A tricky code to break

Reading the long string of four letters repeating in various combinations is proving to be even trickier than scientists thought it would be. At first they believed the genes -- the sequences that control production of the protein building blocks of the body -- would be the only functioning parts of the sequence.

But it turns out there are sequences that control the genes, and perhaps that do even more.

'It now seems that about 5 percent of our genomes are functionally important,' Green said.

'Only a third of that codes for genes. That means that two-thirds of what is functionally important is not (gene) encoding DNA. We don't even know what it looks like so how are we going to find it?'

Green hopes to do so by comparing the genomes of different species.

'This is the idea that you can truly use sequences from multiple genomes and analyze them all at once to try and find the small percent that is shared among all of them,' he said.

'We believe this is going to be a very valuable way to find those sequences that are very important.'

Why? Because nature cannot ditch vital DNA.

'Evolution is about mixing things up. If evolution has hung on to some stretch (of DNA), even if it is repetitive, it is telling us something ... Where did (evolution) find it wasn't safe to change the genome?' he asked.

Green's team, working with researchers at Pennsylvania State University, the University of Washington and the University of California, Santa Cruz, will look at 100 different regions of the genomes of the 13 species.

This is the first region they have analyzed, they reported in Thursday's issue of the journal Nature.

'We will discover new types of functional elements in the coding DNA that we didn't even know existed,' Green said.