New Approaches to Bacterial Taxonomy: Use of Computers

Abstract

An outline of bacterial classification based on numerical taxonomy, genetics and serology is summarized. The main groups are the gram-negative forms; the gram-positive forms; and the actinomycetales. Smaller groups include the spirilla (possibly including some Athiorhodaceae), and Streptobacillus [possibly related to coryne-bacteria (136)]. The gram-negative bacteria are divided into several genera. One corresponds to the Enterobacteriaceae, but including Pasteurella pestis, P. pseudo-tuberculosis and Vibrio comma Proteus and Aeromonas may be intermediate with other families. Another cluster contains Brucella, Neisseria, Actinobacillus (in part), P. multocida, Moraxella, Herellea, Achromobacter and Alcaligenes (in Part.) The Pseudomonadales form a third cluster, including Pseudomonas, some vibrios, perhaps Aero-monas and Photobacterium; its borders are indistinct, and it may con-nect through Xanthomonas with the Rhizobium-Agrobacterium complex and some forms of Arthrobacter Bacillus seems closer to gram-positive cocci. The gram-negative forms than to the typical gram-positive cocci. The gram-positive group includes Staphylococcus, Micrococcus, Sarcina, Streptococcus, Diplococcus, Aerococcus, Listeria, Erysipelo-thrix and somecorynebacteria; internal subdivisions are not clear. The actinomycetales include Nocardia, Streptomyces, Actinomyces, Mycobacterium, Micromonospora, and StreptosporangTum. The internal structure is not clear. There may be no sharp division between bacteria and some blue-green algae. There is a close relationship of rickettsia and pleuropneumonia organisms to eubacteria. Numerical taxonomy is valuable in bacterial classification in producing improved phenetic classifications. In the absence of morphological detail, some such method of handling many characters is essential. Numerical taxonomies will be stable (being based on a representative set of many attributes rather than a few), while the introduction of quantitative methods will, advance taxonomy. Statistical methods will become more sophisticated as distinctions are made between different kinds of taxonomic relationship, and as genetic and evolutionary problems are increasingly studied. Molecular biology will provide data for numberical taxonomy. Such data will need cluster analysis, and will be so numerous that only computers will be able to handle them.