103
Modeling of Biomolecular Systems Interactions, Dynamics, and Allostery Poster Session II
48-POS
Board 1
Similarity-based 3D Modeling of Compound -Protein Complexes
Takeshi Kawabata
, Akira Kinjo, Haruki Nakamura.
Osaka University, Suita, Osaka, Japan.
3D complex structures of proteins and other molecules provide a clue for mechanism of
interaction. It is important to model them in computational methods because only small amount
of 3D complex structures for known interacting molecular pairs are available. Similarity-based
modeling (template-based modeling) of compound-protein complexes can be performed by
similar way to model protein monomers. If the experimental complex 3D data of a similar
compound to the target compound is available, the protein-bound structure of the target
compound can be predicted by aligning the conformation of the target compound on the known
3D structure of the reference. We developed the flexible alignment program
fkcombu
, which
aligns the target compound based on atomic correspondences with the reference compound. The
correspondences are obtained by the MCS (maximum common substructure) of 2D chemical
structures, using our build-up algorithm (Kawabata,
J.Chem.Info.Model.
, 2011,
51
,1775). The
prediction performance was evaluated using many target-reference pairs of superimposed ligand
3D structures on the same protein in the PDB, with different ranges of chemical similarity. We
found that the RMSD between the predicted and correct target conformations significantly
correlates with the chemical similarities between target-reference molecules. Generally speaking,
if the reference and target compounds have more than 70 % chemical similarity, then the average
RMSD of 3D conformations is less than 2.0Å. The source codes of
KCOMBU
package are freely
available. We also developed the server
HOMCOS
(
) to search and
model complex structures. The server separates component molecules of PDB files of complexes
(such as proteins, nucleic acids, small chemical compounds), stores their binding relationships. It
searches these molecules by
BLAST
and our chemical structure comparison program
KCOMBU
.
Based on found similar complexes, simple template-based models of the complex can be
generated.
