Computational Approaches in Peptide and Protein Design: An Overview - Peptide and Protein Design for Biopharmaceutical Applications - page 5

through determining the peptide sequences most suitable for performing
certain biological functions, to de novo predictions of large proteins inter-
acting with one another. In fact, many of the most recent examples are
described and reviewed in later chapters of this topic. Therefore, we do not
feel a real need to review the current state of computational approaches in
detail. Instead, this chapter focuses only on the most general problems of
computational design of proteins and especially peptides, as well as on the
basic techniques required for the practical implementation of design meth-
ods. Our main goal is to introduce the basic elements of the computational
approaches to those interested in peptide and protein design, as well as to
share some of our thoughts and reflections on the current state of the field.
In the first part of the chapter, we discuss computational tools and proce-
dures connected with conformational flexibility of peptides and proteins;
in our opinion, these problems can be satisfactorily resolved only by
applying computational approaches. The second part contains a recent
example of the application of these tools to the specific task of studying
possible complexes between the CXCR4 G-protein-coupled receptor
(GPCR) and its cyclopentapeptide inhibitors [6,7].

2.2 BASICS AND TOOLS

2.2.1 The Importance of Computational Approaches

From a very general point of view, most of the problems of modern peptide
and protein drug design fall into two main areas: structure-based and
target-based design. In the first case, one starts from a parent peptide
('ligand') with no details of the structural information on its specific target
('receptor'). In the second case, structural information on the receptor, at
least on the receptor site that binds the ligand, is available at varying
resolution. In both cases, the goal is to suggest compounds that would
exhibit certain biological qualities (affinity, activity, etc.) as well as or
better than the parent ligand. In structure-based design, the most essential
requirement is to determine the 3D arrangement of the functional groups
of ligand comprising the so-called '3D pharmacophore', which is respon-
sible for ensuring correct interaction between the ligand and the receptor.
In target-based design, one also aims to determine the correct binding
mode of each ligand within the binding site of the common receptor.
However, in the frame of structure-based design, the available experi-
mental methods of structural determination often fail to determine possible
3D pharmacophores characteristic for the interaction of the peptide ligand