t is pretty much impossible to do justice here to the mouse genome paper at #1. More even than the first drafts of the human genome—and in large measure because of the existence of that information—the mouse genome has opened the eyes of biologists to vast deposits of buried genetic treasure. The paper at #1, by the Mouse Genome Sequencing Consortium, points out that "comparative genomics allows one to read evolution’s laboratory notebook," and "to link laboratory notebooks of clinical and basic researchers." Those metaphors may come easily to people directly involved in that kind of work; what about the rest of us?

Paper #1 offers a handy-dandy executive summary of the similarities and differences between the mouse and human genomes, each worthy of an entire article. To focus on one, the multitude of repetitive sequences so often characterized as "junk" DNA contains revelations about the evolution of the two species. The "junk" can be thought of as the fossilized remains of transposable elements, pieces of DNA, often derived from viruses, that can move around the genome, duplicating and inserting themselves here and there. Roughly 46% of the human genome can be recognized as derived from these so-called "transposons." In mouse the figure is about 37.5%. The ratio is close enough to the ratio of genome sizes (2.9Gb vs. 2.5Gb) to tempt scientists to explain the smaller mouse genome in terms of lower transposon activity in mice since their lineage split from the human line about 75 million years ago. In fact transposons have been more active in mice than in humans; the lower figure is a result of higher substitution rates in the mouse, making it harder for researchers to recognize repeats as derived from transposons.

The idea that nucleotide substitutions occur more readily in the mouse has been around since the late 1960s and bedevilled many early attempts to use molecular information as an evolutionary clock. The massive amounts of sequence data now available make it possible to measure substitution rates directly. And the fact that older sequences, which entered the genome before the two lineages split, have perforce been present in both lineages for exactly the same amount of time, deals with problems of phylogeny. Substitution rates in the mouse are about twice as high as in humans. The actual substitution rates, based on a date of 75 million years ago for the split, 2.2 x 10^-9 and 4.5 x 10^-9, are broadly in agreement with previous estimates based on much less data. Note, though, that these rates are per year, not per generation. Human generations are at least 20 times longer than mouse generations, so the actual substitution rate per generation is much higher in humans.

Transposons have often been thought of as enabling evolution, because when they land in a gene they may knock it out, or at least alter its effects, thus creating diversity for selection to act on. It follows that at least to some extent the host genome should fight back, preserving the integrity of genes that really matter. It has. Mouse and human have kept the Hox genes, which regulate early embryonic development, all but spotless. Other regions that are similarly clean in the human are also clean in the mouse, and vice versa. Searching for clean areas in the mouse, the Consortium identified several small regions where repeats are rare. In most, the repeat-rare region is roughly the length of a single gene. It does not seem farfetched to suppose that these genes, presently of unknown function, will prove to be pretty important.

By the same token the host genome should also bury transposons where they will do the least damage, and so it proves. Mouse and human genomes both tend to have twice as many transposons near the ends of the chromosomes. This applies to younger transposons too, those that entered the lineages after they had separated. So the ends of chromosomes seem to be safe hiding places.

What of the notion that all this repetitive DNA is in fact junk? The jury is still very much out, but the much more active behavior of transposons in mice should make it possible to see whether, for example, the presence of a transposon next to a gene alters the expression of the gene enough to make it more mutable. Whole-genome shotgun sequencing—the technique pioneered by Celera to create its draft of the human genome and used by the Mouse Consortium to draft the mouse genome—is notoriously poor at picking up repeated sequences. But the Mouse Consortium plans to use "proper" sequencing to complete the task, which should enable researchers to further investigate the usefulness or otherwise of all that junk.

There is more; much, much more. Having the two genomes side by side is going to make it considerably easier to create and study accurate mouse models of human disease, for one. And this discussion has said nothing at all about the equally fascinating paper at #7, an entirely new approach to making sense of the mouse genome.