The Database of Macromolecular Motions: new features added at the decade mark

Abstract

The database of molecular motions, MolMovDB (http://molmovdb.org), has been in existence for the past decade. It classifies macromolecular motions and provides tools to interpolate between two conformations (the Morph Server) and predict possible motions in a single structure. In 2005, we expanded the services offered on MolMovDB. In particular, we further developed the Morph Server to produce improved interpolations between two submitted structures. We added support for multiple chains to the original adiabatic mapping interpolation, allowing the analysis of subunit motions. We also added the option of using FRODA interpolation, which allows for more complex pathways, potentially overcoming steric barriers. We added an interface to a hinge prediction service, which acts on single structures and predicts likely residue points for flexibility. We developed tools to relate such points of flexibility in a structure to particular key residue positions, i.e. active sites or highly conserved positions. Lastly, we began relating our motion classification scheme to function using descriptions from the Gene Ontology Consortium.

Figure 1

The motion of an ethane molecule as determined by geometric simulation in FRODA. (a) Initial atomic positions; (b) ghost templates; (c) random atomic displacement; (d) fitting of ghost templates to atoms; (e) refitting of atoms to ghost templates; (f and g) further iterations of (d and e); (h) until a new conformer is found¹³.

Figure 2

The new morph page. The previous page, now called morph-classic.cgi, can still be accessed by a link. Features: (a) a page with links to PDB entries with >99% sequence homology; (b) highlight active sites from the CSA database, if entries exist in any close homologs; (c) Torsion angle plots can be useful in guiding your hinge selection efforts; (d) if you wish to contribute to our hinge research, use the arrow buttons to manually select up to three hinges by visual inspection; (e) Submit your hinge selection, plus any comments. Comments appear in our public bulletin board; (f) if our FlexOracle hinge prediction program has been run on the first frame of the morph, the energy versus residue number plot can be viewed here; (g) the 10 best (lowest energy, in ascending order of energy) hinges can be highlighted in the viewer.

Figure 3

Example plots of the distribution of maximum Cα displacement of the mobile component (second core) in structures annotated with GO terms (‘DNA binding’, 17 morphs—top; ‘metabolism’,17 morphs—centre), compared with the reference morph dataset (200 canonical morphs—bottom). Dark bars indicate morphs whose second core max. Cα displacement falls below the median, while light bars indicate those above.