基本HTML版本
Intl. J. of Mol. Evol. and Biodivers. 2012, Vol. 2, No.1, 1-7
http://ijmeb.sophiapublisher.com
3
Lactococcus lactis
, former two belonged to the same
subspecies. On one hand, the Euclidean distances of
these three strains were calculated by formula (3). The
results revealed that the Euclidean distance between
L.
lac
_MG and
L. lac
_SK was 0.0438, while that of
L.
lac
_MG and
L.lac
_I1, which belongs to the different
subspecies, was 0.0624 (Table 1). The difference
between taxa of “the same subspecies” and taxa of
“the different subspecies” was just 0.0186. On the
other hand, the difference of JHD between taxa of “the
same subspecies” and taxa of “the different
subspecies” was 51.6914 calculated by formula (2)
(Table 2). It suggested that the JHD had a higher
discrimination for the closely related organisms than
the Euclidean distance did.
In addition, comparing
L. lac
_MG with Pediococcus
pentosaceus ATCC 25745 (abbreviated as
P. pen
_AT)
in the same order, with
Staphylococcus aureus subsp.
aureus
USA300_FPR3757 (abbreviated as
S. aur
_US)
in the same class and with
Alkaliphilus
metalliredigens
QYMF (abbreviated as
A. met
_QY) in
the same phylum, the Euclidean distances were 0.3209,
0.3918 and 0.3823, respectively, and the mean,
standard deviation and coefficient of variation for the
Euclidean distances of this group were 0.3650, 0.0385
and 0.1054, respectively.
It indicated that Euclidean distance couldn’t reflect the
evolutionary relationship for taxon above order
because the difference within class was greater than
across classes. However, the JHD of
L.lac
_MG and
P.pen
_AT,
L.lac
_MG and
S.aur
_US,
L.lac
_MG and
A.met
_Qy were 1090.6251, 1291.8410 and 2305.8140,
respectively, and the mean, standard deviation and
coefficient of variation for JHD of this group were
1562.7600, 651.3209 and 0.4168. The difference
between them increased with the rise of their taxonomic
level. Similarly results could be obtained for 6 strains of
Staphylococcus aureus subsp. aureus
and 6 strains of
Streptococcus pyogenes
(Supplement 3). So, compared
with the Euclidean distance, the JHD has more effective
in estimation of phylogenetic relationship.
1.3 The joint histogram divergence could be a
complement for the construction of phylogeny tree
Take the 5 different strains of
Staphylococcus
for
example, in which former 4 strains belonged to
Staphylococcus aureus
and the last one belonged to
Staphylococcus saprophyticus
. The phylogenetic tree
based on 16S rRNA gene sequence revealed that the 4
strains of
Staphylococcus aureus
could cluster
together, but their evolutionary relationship was not
clear (Figure 2a). It indicated that the evolutionary
relationship of organisms could not be determined just
based on sequences alignment of single gene. When
we then reconstructed phylogenetic tree based on joint
histogram divergence matrix, the 4 strains belonged to
the same species could not only cluster together but
also be classified clearly (Figure 2b). The similar
conclusion could be obtained for the 6 strains of
Streptococcus pyogenes by the same way (Supplement
4). This exciting result showed the JHD might be a
helpful complement while the single gene cannot
determine the evolutionary relationship clearly.
Figure 2 The phylogeny tree of 5 strains of Staphylococcus
Note: (a) represents the phylogeny tree based on 16S rRNA
gene constructed by neighbor-joint method; (b) represents the
phylogeny tree based on joint histogram distance matrix
2 Discussions
The novel method that was applied to deduce the
evolutionary relationship by oligonucleotide transition
probability matrix could take account of the
information of whole genome and should reflect the
evolutionary history of organisms more truly in theory.
Here, the diversity between any two of microbial
organisms was compared by image registration
technology, measured quantitatively by its JHD. As
we know that JHD is based on the statistics for
grayscale each pixel of two color matrixes, it has an