TABLE 3.

Insights into the relationships of genotype and phenotype^a

Type of relationship	First strain	Second strain	No. of bp differences	Reason for this example
Same genotype but different phenotypes	Mycobacterium tuberculosis ATCC 27294	M. bovis ATCC 19210 or M. africanum ATCC 25420	0	Although given names that appear to be species, the three strains are now designated subspecies and show small or no genetic difference, but there is a definite phenotypic and pathological difference. See reference 43 for other examples within the mycobacteria. This rare situation is a limitation to the use of sequencing as gold standard.
Similar genotype but different phenotypes	E. coli ATCC 11775	Shigella dysenteriae ATCC 13313	3	Two high-level pathogens are genotypically close enough to be considered the same species but have kept clinically and historically important separate names.
	Streptococcus pneumoniae ATCC 33400	S. mitis ATCC 49456	3	A high-level pathogen and a commensal are genotypically very similar.
	Streptococcus bovis	S. equinus	2	Very close genogroups which are difficult to distinguish phenotypically in the clinical laboratory. However, their reported difference in lactose reaction and different niches may justify the different names.
Similar phenotypes but different genotype	Nocardia asteroides ATCC 19247	N. farcinica ATCC 3318	13	Separate species which are relatively difficult to distinguish phenotypically are easy to distinguish by sequence.
	S. bovis ATCC 33317T	S. bovis ATCC 43143	13	Separate genogroups which are difficult to distinguish phenotypically are easy to distinguish by genotype (11).
Too distant to be the same species	Enterobacter (Pantoea. agglomerans (bg1)	E. (Pantoea. agglomerans (bg2)	27	Strains that were originally thought to be biogroups within the same species are genotypically distant enough to be considered separate genera.
Too distant to be the same genus	Clostridium tetani ATCC 19406	Clostridium innocuum ATCC 14501	About 104	Although these two organisms have been given the same genus name, the large difference (20%) means that at least one has been taxonomically misplaced. Species in the same genera should not differ by more than about 5 to 6%.
Too close to have different names	Mycobacterium duvalii	“M. valentiae”	0	Sometimes names get into the literature without full justification.
Too close to be three different genera	Enterobacter cloacae ATCC 13047	Leclercia (Enterobacter. adecarboxylata ATCC 23216	1-2	These three organisms, which have been placed in different genera, are genotypically close enough to be considered the same species.
	E. cloacae ATCC 13047	Citrobacter werkmanii	6
Subspecies	Streptococcus dysgalactiae subsp. dysgalactiae	Streptococcus dysgalactiae subsp. equisimilis	14/500 and 16/1,500	The difference between these two “subspecies” is greater than that between the genera above.
	Streptococcus equi subsp. equi	Streptococcus equi subsp. zooepidemicus	1/500 and 1/1,500	The only difference between the two subspecies was at bp 204.
	Staphylococcus cohnii subsp. cohnii	Staphylococcus cohnii subsp. urealyticum	1/500 and 3/1,500	The difference of 0.33% was the same calculated using either the 500-bp or the 1,500-bp sequence.
Other points	Streptococcus mitis ATCC 49456	Clinical strains identified as S. mitis	3-10	Lower-level pathogens tend to have more heterogeneous genotypes for a phenotypically derived name, presumably because of poor phenotypic discrimination.
	S. intermedius ATCC 37335	S. constellatus ATCC 27823	9	Although they are sometime considered together as the S. milleri group, S. intermedius is much more closely related to S. constellatus than to S. anginosus.
	S. intermedius ATCC 37335	S. anginosus ATCC 33397	33
	S. anginosus ATCC 33397	S. anginosus, urinary tract strains	6	There is almost as much variability among the S. anginosus clades or subspecies as between the genera Enterobacter (cloacae) and Citrobacter (werkmanii) above

^aExamples of base pair differences per 500 bp sequenced between the two strains listed and the significance of this result.