Biology Reference
In-Depth Information
Another example of structural accordance with the presented model is the
transmembrane protein discussed in Zobnina and Roterman ( 2009 ) , where the
central hypothesis of the “fuzzy oil drop” model is validated in the context of
interaction between proteins and cellular membranes.
6.2.2
HADDOCK - High Ambiguity Driven Biomolecular
DOCKing Based on Biochemical and/or Biophysical
Information
HADDOCK implements algorithms for systematic search on a grid (Dominguez
et al. 2003 ; de Vries et al. 2007 ). Its key innovation lies in a heuristic approach
to experimental data, including NMR data such as residual dipolar couplings and
relaxation (van Dijk et al. 2005a, b, c ) , sequence conservation, mutagenesis,
epitope mapping, H-D exchange or crosslinking experiments to provide distance
restraints, diffusion anisotropy (van Dijk et al. 2006a, b ) , solvated docking (van
Dijk and Bonvin 2006 ) and flexible protein-DNA docking (van Dijk et al. 2006b ) .
If no experimental cues are available, docking is guided by randomly selected
patches of tentatively active residues and restrained by molecular center-of-mass
criteria.
Prior to running HADDOCK a set of Ambiguous Interaction Restraints (AIRs)
has to be generated. This procedure involves distinguishing “active” and “passive”
residues. Active residues are those which interact with the target protein while
remaining in contact with water. Passive residues are also exposed to water and lie
in direct proximity to active ones. The cutoff criterion is not explicit; rather, it
acknowledges the structures of important functional groups. Determination is based
on NMR data and requires end-user assessment.
HADDOCK also enables identification of protein-protein interaction candidates
on the basis of the so-called conservative sequences. This process necessitates further
division of residues into “active” and “passive” sets (de Vries et al. 2006 ) . If none
of the presented approaches is feasible (e.g. due to the lack of NMR input),
identification can proceed by way of modeling solvent-accessible areas.
HADDOCK computations have proven quite successful in several editions of the
CAPRI challenge (van Dijk et al. 2005a, b ; de Vries et al. 2007 ) .
6.2.3
RosettaDock
RosettaDock works as an extension of the Rosetta structure prediction package.
Similarly to other docking algorithms, the first phase of the search involves sam-
pling the rigid-body conformational space with side chains represented by single
pseudo-atoms. Docking conformations may be refined in the second (full-atom)
phase via small-scale perturbations of the complex and side chain optimization,
employing rotamer packing and continuous minimization in order to avoid entrapment
Search WWH ::




Custom Search