If you want to appreciate the diversity of life on earth, you will need a microscope. There are about 5,400 species of mammals on the planet, but just a spoonful of soil may contain twice as many species of microbes. They can dwell in habitats where so-called higher life forms like us would quickly die, including acid-drenched mines and Antarctic deserts. By one rough estimate, there may be, all told, 150 million species of microbes.

"Microbes represent the vast majority of organisms on earth," said Hans-Peter Klenk, a microbiologist for the German Collection of Micro-organisms and Cell Cultures, a government microbiology research centre.

Yet scientists still know very little about our microbial planet.

The genomes of only about 1,000 species of microbes have been sequenced. That leaves 99.99999 per cent to go. Making matters worse, the genomes scientists have sequenced so far are clustered together in groups of closely related species, leaving vast stretches of the microbial tree of life virtually unexplored. It would be as if all we knew about the animal kingdom were based entirely on a stuffed ferret and a pickled tarantula.

To shed light on this enormous stretch of biological darkness, the Joint Genome Institute at the US Energy Department has started what it calls a "genomic encyclopedia." It is filling the encyclopedia with the genomes of microbes from remote reaches of the tree of life.

In the December 24 issue of Nature, Klenk and his colleagues present their first analysis of the encyclopedia, based on the first 56 species they have sequenced.

Using this new evolution-based approach, the scientists have discovered many kinds of genes, some of which may prove a boon to the biotechnology industry.

"The encyclopedia is guaranteed to yield new things," said Norman Pace, a University of Colorado microbiologist who was not involved in the study. "We humans haven't even scratched the surface of natural microbial diversity."

When scientists first began to pick microbes for genome sequencing in the 1990s, they favoured species they had been studying for years, like E coli. As the technology improved and costs fell, they moved from these microbial lab rats to species that were important to humans for one reason or another, like those causing diseases.

It gradually became clear that this approach neglected most of the diversity of microbes. A number of microbiologists began to argue that a broader survey of the microbial world would bring many new insights. Comparing a gene in many different species can often help scientists figure out what a gene does in the first place, for example.

"It's been blatantly obvious that this should be done," said Jonathan Eisen, an evolutionary biologist at the University of California, Davis, and the lead author of the Nature paper.

Taking advantage of the falling cost of DNA sequencing, Eisen, Klenk and their colleagues at the genome institute established the Genomic Encyclopedia of Bacteria and Archaea. (Bacteria and archaea are two of the major branches of the tree of life. The third branch contains eukaryotes, which includes animals, fungi, plants and protozoa.) The scientists selected 200 species to analyse.

Once the scientists had sequenced 56 genomes, they decided to see whether their approach was paying off. For part of their analysis, they tallied up how many new genes they had found: tens of thousands. But more importantly, they found 1,768 new gene families -- sets of genes that share a common ancestor.

One gene, from a microbe that lives in salt flats, encodes an enzyme that can cut up cellulose in the presence of lots of salt.

This type of enzyme might be useful for extracting biofuels from plants. "The biotech industry loves these things," he said.

The encyclopedia is also upending some basic rules of biology. All eukaryotes, ourselves included, have Lego-like skeletons in our cells made from a molecule called actin. "We have this rule that actin is in eukaryotes," Eisen said, "and it's not in bacteria and archaea." Now he and his colleagues have found a gene for actin in a species of marine bacteria. "It's wildly cool," he said.

The genomic encyclopedia includes not only the raw data on genes, but also predictions about what those genes are actually for.

The genome institute has set up an "adopt a genome programme" to enlist the help of college students. Undergraduate microbiology students can pick a species from the encyclopedia and analyse its biology.

"We've made a dent," Eisen said, "but it's a small dent."

International Herald Tribune