Biology Reference
In-Depth Information
each amino acid has a 1% chance of mutation. For instance, multiplying
PAM by itself 20 times—often the result is called PAM20—yields a ma-
trix in which each amino acid has a 20% chance of mutating. As Day-
hoff was aware, using matrices such as PAM250 can be extremely help-
ful in detecting distant evolutionary relationships between proteins.
Dayhoff's larger aim was to use such models of mutation to explore
the evolutionary relationships among all proteins. From the fi fth edi-
tion onward (1973), the
Atlas
was organized using the concept of pro-
tein “superfamilies,” Dayhoff's second major analytical contribution.
Families of proteins were already well recognized and easily determined
through simple measurements of sequence similarity. The sensitivity of
Dayhoff's methods of comparison (using the PAMs), however, allowed
her to sort proteins into larger groups, organized according to common
lines of descent.
22
Such classifi cations were not merely an organizational
convenience—they provided theoretical insight into the process of evo-
lution. The ultimate aim of the NBRF's sequence collection work was
this kind of conclusion:
In examining superfamilies, one is struck by the highly con-
servative nature of the evolutionary process at the molecular
level. Protein structures persist through species divergences and
through gene duplications within organisms. There is a gradual
accumulation of change, including deletions and insertions as
well as point mutations, until the similarity of two protein se-
quences may no longer be detectable, even though they may be
connected by a continuum of small changes.
23
The superfamily concept was both a tool of classifi cation and a biologi-
cal theory. It was a way of conceptualizing the relationships among the
entities that made up living things and of making sense of their history.
In an article published in
Scientifi c American
in 1969, Dayhoff outlined
some of the conclusions of her work on the classifi cation and history of
life: “The body of data available in protein sequences,” she argued, “is
something fundamentally new in biology and biochemistry, unprece-
dented in quantity and in concentrated information content and in con-
ceptual simplicity . . . because of our interest in the theoretical aspects
of protein structure our group at the National Biomedical Research
Foundation has long maintained a collection of known sequences. . . .
In addition to the sequences, we include in the
Atlas
theoretical infer-
ences and the results of computer-aided analyses that illuminate such
inferences.”
24
