CoCo (Complementary Coordinates) is a tool which may be used to enrich an ensemble of structures, presented in the form of a multi-model PDB file.

Background

NMR structures are typically deposited in the PDB in the form of an ensemble of structures. Even though each of the models in such an ensemble satisfies the experimental data and is equally valid, the limited number of structures that are typically deposited cannot completely encompass the structural diversity allowed by the observed NMR data, even for relatively rigid molecules. However, they can be chosen to try and maximise its representation.

CoCo, which is based on principal component analysis, analyses the distribution of an ensemble of structures in conformational space, and generates a new ensemble that fills gaps in the distribution. These new structures are not guaranteed to be valid members of the ensemble, but should be treated as possible, approximate, new solutions for refinement against the original data.

CoCo is freely available from this page by uploading your PDB file onto our server. The file should contain between 10 and 25 distinct models for your structure. CoCo will analyse this ensemble, report a few characteristics, and return a new multi-model PDB file with an equal number of new, suggested, structures. The method is very rapid - analysis of a 25 member ensemble will typically take 1-2 minutes. Though developed with protein NMR data in mind, the method is quite general – the initial structures do not have to come from NMR data, and can be of nucleic acids, carbohydrates, etc.

For full details of the method, see:

Laughton C.A., Orozco M. and Vranken W., COCO: A simple tool to enrich the representation of conformational variability in NMR structures, PROTEINS, 75, 206-216 (2009)

Any published work which utilizes CoCo should include this reference.

Method overview

The outline of the CoCo method is as follows:

  1. The submitted ensemble is subjected to Principal Component Analysis and the position of each structure in the PC1/PC2/PC3 subspace identified.
  2. New points are placed to fill gaps in the distribution.
  3. The new points are converted back into new structures, which are returned to the user.