Next, they tried the FFP method on 518 genomes, the bulk of them bacteria and Archaea, but also six eukaryotes of varying complexity and two random sequences. The eurkaryotic genomes used were as much as 1,000 times longer than the bacterial and Archaeal genomes. Because most of the bacterial and Archaeal genomes code for genes, as opposed to very little of the genomes of higher eukaryotes, the researchers used a different alphabet and vocabulary for the FFP method: short strings of amino acids, the building blocks of proteins, with a 20-word alphabet representing the 20 possible amino acids.
"The question is: Can we then group all living organisms based on the whole proteome, that is, the assembly of all proteins, instead of using just a selection of a small set of proteins, which is equivalent to using a small set of genes?" said Kim.
The researchers found that the FFP method clearly segregates whole proteomes of all bacteria, archaea, eukaryotes and random sequences into separate groups or domains. Most of the phylum groups within each domain and class groups in each phylum also were well segregated, with some interesting discrepancies compared to the currently accepted groupings.
In most of the cases where the FFP grouping disagreed with an accepted phylogenetic grouping, the problem organism had been the subject of debate among biologists because of conflicting conclusions from genetics, behavior and morphology, Kim sa
|Contact: Robert Sanders|
University of California - Berkeley