Appeal 2007-0537
Application 10/102,902
Further . . . the total number of conformations in a
molecule may be significantly more than the conformations
in a fragment pair (Application at page 22, lines 7-20). Thus,
the claimed invention recognizes that using fragment pair
representation may require significantly smaller storage than
a brute force enumeration (Application at page 22, lines 20-
24). Nowhere is this recognized by Cornileseu.
(Reply Br. 1-2 (bolding added for emphasis).)
Appellants’ arguments are unavailing, particularly since they do not
focus on the claim language or limit the scope of the present claims. (FF 8;
see also the permissive language in the passages quoted above.) In any case,
Cornilescu does recognize the value of using select features of adjacent
amino acid residues, i.e., secondary chemical shifts and amino acid
sequence. (FFs 10, 11, 14-18.) This technique takes into account the effect
of surrounding residues on chemical shifts and “avoids the need to search”
every characteristic of the adjacent residues in his database to identify other
residues with similar backbone torsion angles.
As found by the Examiner,
chemical shifts of the backbone atoms are functions of the
molecular properties of the neighboring amino acids as well
as the amino acid itself. . . . Since the values of the chemical
shifts change between their values in isolated amino acids to
their values in proteins, the search with these modified values
of chemical shifts are actually searches of fragment pairs of
amino acids and not merely a single residue.
(Answer 17.)
This is not a case in which Appellants invented a new computer, or a
new search method, or even a new method of molecular superposition (i.e.,
alignment). (FFs 22-24.) Instead, Appellants are using systems and
techniques well known in the art to identify molecules with similar structure
13
Page: Previous 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Next
Last modified: September 9, 2013