Biology Reference
In-Depth Information
Chapter 9
Protein Structure Modeling
and Docking at the Swiss Institute
of Bioinformatics
Torsten Schwede and Manuel C. Peitsch
1. Introduction
Knowledge of the three-dimensional (3D) structures of proteins and
their interactions with other molecules provides invaluable insights into
the molecular basis of their functions and a rational basis for empirical
functional analysis through site-directed mutagenesis, mapping of
disease-related mutations, or the structure-based design of specific
inhibitors.
1
While structure determination methods such as X-ray crystal-
lography,
2
high-resolution electron microscopy,
3
and nuclear magnetic
resonance (NMR) spectroscopy
4
have greatly progressed, they are still
expensive, time-consuming, and not always applicable. Currently, about
51 000 experimental protein structures have been released by the Protein
Data Bank (PDB)
5
; these structures correspond to approximately 18 000
different proteins (sharing less than 90% sequence identity among each
other). However, the number of structurally characterized proteins is
small compared to the 400 000 annotated and curated protein sequences
in the Swiss-Prot section of UniProtKB
6
(http://www.expasy.org/sprot/).
This number appears even smaller when compared to the 5.9 million
known protein sequences in the complete UniProtKB (release June
2008). Even after removing the highly redundant sequences from this
database (above), the remaining 4.1 million sequences exceed the number
219
