Ask someone to think of an animal, and chances are they'll come up with one of our relatives among the mammals. A few people might go further afield and mention other vertebrates, like birds and fish. But these barely scratch the surface of animal diversity, with things like cephalopods, insects, and echinoderms all having distinct features.

And that's before you get to the really weird stuff, like the radially symmetric Cnidarians, or the sponges that lack muscle and nerve cells. Or the comb jellies, which move themselves around by spinning lots of thread-like cilia. Or the truly bizarre placozoans, disk-like creatures that have two sides but no interior and digest things on their surface.

For people who tend to think that evolution involves adding ever-greater complexity to organisms, it's tempting to imagine that the animal family tree came about by progressively adding more stuff, like nerve cells and muscles. But there has been a steady flow of genetic studies that hint that there are two separate lineages that ended up with nerve cells. The results of these studies were a bit dependent upon the genes and species chosen for the analysis. But a new study that's not as dependent upon individual genes now firmly places sponges as more closely related to humans than some other animals with a nervous system.

Chromosomal reorgs

Most of the early studies in this area involved identifying related genes that are present in all animals and figuring out how those genes are related. Presumably, the organisms themselves are related in the same way. That can be very informative in many situations, but the analysis tends to get confused when lots of species branch off in a short amount of time, or when individual genes change a lot due to evolutionary pressures. So, the exact answer you get can sometimes depend on what genes you choose to look at.

The new study tries to avoid the confusion by looking at how genes are arranged on chromosomes. It turns out that individual genes tend to stay in the same place on a chromosome for long periods of time; it's estimated that it takes 40 million years for just one percent of the genes in a typical animal genome to move to a new chromosome. So the odds are that if four genes are next to each other now, then they were next to each other in the ancestors of today's mammals that had to avoid being eaten by dinosaurs.

That doesn't mean that those ancestors had exactly the same number and arrangement of chromosomes. Large-scale rearrangements happen, like fusion or splitting of chromosomes, or swapping a big chunk of one to another. But those large rearrangements keep almost all of the nearby genes next to each other, even if the whole group ends up on a different chromosome (swaps can involve as little as one break in a DNA molecule).

That means that breaking up the linear arrangement of a group of genes—technically called synteny—is pretty rare in an animal's evolutionary history. And, by tracking changes in the arrangement of genes across different species, we can figure out where in an organism's past groups of genes got broken up, and which other species inherited the same rearrangement. And that can tell us which organisms are more closely related to us.

Tracking rearrangements

To do this sort of analysis, you need to know how genes are arranged on chromosomes. We've only recently developed technology that allows us to sequence very long pieces of DNA—often tens of thousands of bases in a stretch—which makes piecing together chromosomes much easier. The researchers relied on many animal genomes where this had been done and completed a few of their own for the study. In addition, they reconstructed the chromosomes of single-celled organisms that are thought to be closely related to animals to provide a baseline for the starting arrangements.

The origin of animals is thought to have occurred roughly 800 million years ago. So, although breaking up clusters of genes is rare, that's enough time for it to have happened across much of the genome. The researchers were only able to identify a bit under 300 genes that were in clusters that extended back to the single-celled relatives of animals, with the biggest cluster including 29 genes. When the researchers ran 10 million simulations that randomly broke up genes at the rates expected over 800 million years, they never wound up with a cluster as large as eight genes, so most of these are likely to be real ancestral states.

By tracing the rearrangements, the researchers were able to identify eight rearrangements that were shared by animals with left and right sides like us vertebrates, and things like jellyfish (Cnidaria) and sponges (Porifera). None of these showed up in comb jellies (Ctenophora). Again, they performed 100 million random simulations and never saw this pattern of inheritance, so it seems to be real.

This means that animals like us vertebrates, along with everything else that has a left and right side, is more closely related to sponges than we are to comb jellies. That's despite the fact that sponges have no muscles or nervous system, while comb jellies share both of them with us.