A big problem is that only about one in 100 microbes can be cultured sufficiently to extract its genome. Even if a microbial genome is known, however, this still doesn't tell researchers how it interacts with other microbes in its environment.
Banfield and other researchers have been looking at a more daring approach - sequencing the whole community at once, a technique Banfield prefers to call "community genomics." That's like surveying the species in the African veldt by grinding up lions, zebras, elephants and an unknown number of other animals, cutting the genes into tiny pieces, and trying to sort them into distinct genomes.
But it works, Banfield said, "at least with the small number of distinct organisms in this community."
"The magic of the whole thing is that, because of speciation, these organisms are different enough that their genomes are easy to tell apart," Banfield said. The technique would allow researchers to sequence the genomes of microbes that cannot be raised in isolated cultures.
Banfield's group and the group at the JGI reassembled the genomes, each containing about 2,000 genes, using different software programs to arrive at composite genomes. Two of the draft genomes - for a Leptospirillum group II bacterium and a Ferroplasma type II microbe from the ancient group known as Archaea - are now about 97 percent complete, with a few gaps. The genome of one of the six microbes in the community, Ferroplasma acidarmanus (a type I Ferroplasma), had been sequenced earlier by JGI and Banfield's group and was a good control during the sequencing and assembly process.
The other genomes were highly fragmented, but identifiable as microbes from Leptospirillum group III, Ferroplasma type I and a G-plasma microbe.
"Despite the highly fragmented nature of three of the genomes, we had enough coverage and many, many genes to get an idea of what the organisms do in the environment," Tyso
'"/>
Contact: Robert Sanders
rls@pa.urel.berkeley.edu
510-643-6998
University of California - Berkeley
1-Feb-2004