Biomedical Engineering Reference
In-Depth Information
Substitution Matrices
Protein structure and function are surprisingly resistant to polypeptide substitution, to the degree
that the substitutions don't alter the chemistry of the protein. Substitutions are common over large
expanses of time and from one species to the next. In many cases, the substitution of polypeptides
through evolution can be predicted. In this way, a matrix of likely polypeptide substitutions can be
constructed. As in a dot matrix analysis, the amino acids are listed across the top and side of a
matrix, typically using the amino acid code letters listed in
Table 8-3
. At each intersection, the matrix
is filled with a score that reflects how often one polypeptide would have been paired with the other in
an alignment of related protein sequences. An underlying assumption is that this association is
symmetrical, in that either polypeptide can be substituted for the other.
Two popular substitution matrices are Percent Accepted Mutation (PAM) and Blocks Amino Acid
Substitution Matrix (Blosum), examples of which are shown in
Figures 8-5
and
8-6
, respectively.
Unlike dot matrix analysis, these matrices are static. Furthermore, these matrices aren't mere
mathematical constructs designed simply to facilitate computational sequence alignment, but they
reflect the biology of the molecules represented by the sequences. For example, each PAM matrix in
the series is named after the level of change assumed by the matrix. For example, the commonly
used PAM-250, shown in
Figure 8-5
, assumes a 250-percent change in the probability matrix.
Table 8-3. Amino Acid Code Letters. Courtesy NCBI.
Code
Meaning
A
Alanine
B
Aspartate or Asparagine
C
Cysteine
D
Aspartate
E
Glutamic Acid
F
Phenylalanine
G
Glycine
H
Histidine
I
Isoleucine
K
Lysine
L
Leucine
M
Methionine
N
Asparagine
P
Proline
Q
Glutamine
R
Arginine
S
Serine
T
Threonine
