Look at the tree. One of its branches is extracted.
It is supported by 29 positions - that's
written on the scheme.
Let's consider some positions. E. g, in the position number 107 phenylalanine
in sequences of the marked groups is placed opposite lysine in all other sequences (out of branch).
It's an ideal case: alignment is divided into two clear groups, this
division is the same with those, given by the branch.
More widespread is the case when positions consists of three or
more amino acids. But on the one side of the branch
(usually - inside the branch) one amino acid is placed. Such are
positions number 5 (leucine opposite isoleucine, valine
acid opposite arginine, glycine and serine),
opposite alanine, phenylalanine and serine),
opposite leucine, histidine and tyrosine),
opposite isoleucine, leucine and valine),
opposite isoleucine, leucine and proline),
opposite glutamine, glutamic acid and valine),
opposite tyrosine and gap) and 155
(valine opposite alanine, isoleucine
It's worth mentioning, that positions 30,
51, 111 and
137 are better that other ones -
because two groups are presented by principially different acids
(in position 5, for comparison,
all aminos - isoleucine, leucine, methyonine and valine -
are rather similar in their physico-chemical properties).
Positions of the third group (6, 7, 26, 38 - 42, 44, 60, 63, 68, 84, 91, 96, 106, 143, 154 and
158) are rather
non-homogeneous. Though two classes of aminoacids (those which are
present in the branch and these which present out of it) don't
intersect, they are rather wide for clear distinquishing the branch.
That's why it's no use paying attention on them.