DNA extracted from the oral and fecal samples was used
as a template for amplification of a portion of the 16S
rRNA gene by PCR, using primers specific for bacterial
16S rRNA. The amplification products were inserted into
plasmid vectors to generate libraries from which individual
clones were randomly selected for DNA sequence
determination. More than 200 clones were sequenced in
each study, with 400–700 bases from each used in subsequent
analyses (Table 1). Sequences greater than 98%
identical were considered as a single ‘phylotype’ likely to
be derived from a single species [6]. Both groups found a
considerably larger number of phylotypes — 59 oral and
82 fecal — than had ever been isolated from a single
person by cultivation. Interested readers should consult
the papers [2,3] for further discussion of the criteria used
to define phylotypes (Table 1), and the rationales behind
them. The consensus among microbial taxonomists is
that a bacterial species should be defined in genetic
terms as organisms sharing a minimum of 70% overall
similarity in genomic DNA sequences, usually as measured
by DNA:DNA reassociation [12]. This generally
corresponds to differences of at most 2–3% in 16S rRNA
sequences between isolates of the same species [12].
(Bacteria often contain multiple rRNA operons in their
genome, but these copies typically differ in sequence by
less than 1% [2].)