MolMovDB:Database of molecular

movements

Anastasia Kurdia

Motivation

Motion is closely related to the way a structure fulfills a particular function, and protein motion is involved in wide variety of basic protein functions

Studying motions provides an insight into function

Motivation

Rapidly increasing amount of data characterizing protein structure (X-ray crystallography, NMR, computational data)

Classification of motions helps in predicting protein function

Motivation

• Need for data storage and management

• Need for an efficient mechanism to relate the structures one to another, beyond classification based on sequence or structural alignment.

MolMov database

• Storing motions• Obtaining motions• Classification of

motions• Standardized

statistics and quantitative measures

• Unified nomenclature for protein motions

Movies as database entries

Custom software tool to perform service tasks – morph server

Protein motions

Manually maintained descriptions of conformational changes in hundreds of distinct proteins, with references and movie links.

Possible ways a motion can be classified

[D-s-2] Known Domain Motion, Shear Mechanism [D-h-2] Known Domain Motion, Hinge Mechanism [D-?-2] Known Domain Motion, Unclassifiable Mechanism [D-n-2] Known Domain Motion, Neither Hinge nor Shear Mechanism [D-f-2] Known Domain Motion, Partial Refolding of Structure [D-s-1] Suspected Domain Motion, Shear Mechanism [D-h-1] Suspected Domain Motion, Hinge Mechanism [D-?-1] Suspected Domain Motion, Unclassifiable Mechanism [D-n-1] Suspected Domain Motion, Neither Hinge nor Shear Mechanism [F-s-2] Known Fragment Motion, Shear Mechanism [F-h-2] Known Fragment Motion, Hinge Mechanism [F-?-2] Known Fragment Motion, Unclassifiable [F-n-2] Known Fragment Motion, Neither Hinge nor Shear Mechanism [F-s-1] Suspected Fragment Motion, Shear Mechanism [F-h-1] Suspected Fragment Motion, Hinge Mechanism [F-?-1] Suspected Fragment Motion, Unclassifiable [F-n-1] Suspected Fragment Motion, Neither Hinge nor Shear Mechanism [S-a-2] Known Subunit Motion, Involving Allostery [S-n-2] Known Subunit Motion, Not Involving Allostery [S-a-1] Suspected Subunit Motion, Involving Allostery [S-n-1] Suspected Subunit Motion, Not Involving Allostery [F----] Known, Notably Motionless Proteins (fragment) [D----] Known, Notably Motionless Proteins (domain) [S----] Known, Notably Motionless Proteins (subunit) [C----] Complex Motion [N-R-1] Suspected RNA Motion [N-R-2] Known RNA Motion [N-D-1] Suspected DNA Motion [N-D-2] Known DNA Motion

• http://www.molmovdb.org/cgi-bin/motion.cgi?ID=calbind

• http://www.molmovdb.org/cgi-bin/motion.cgi?ID=calbind&graphics=1

Sample entry in Protein motions

Movies

Thousands of morphs of transitions between PDB files, viewable through a Java applet or as MPEG or GIF movies. Most of these are submissions to the Morph Server by database users

• http://www.molmovdb.org/cgi-bin/morph-classic.cgi?ID=34318-16829

• http://www.molmovdb.org/cgi-bin/morph.cgi?ID=63984-22780

Sample entries in Movies

Structural domain

A structural domain of a protein is a self-stabilizing structural element that often folds independently of the rest of the protein chain. Many domains are not unique to the proteins produced by one gene or one gene family but instead appear in a variety of proteins

Hinge vs. shear motion

Hinge motion is characterized by large changes in main-chain dihedral angles occurring at a localized region (a hinge). Hinge motions are similar to rotations around an articulated joint and therefore can be very large. They usually involve a small number of residues, since even one bond can provide the required rotational freedom.

• Shear motions are very limited and involve large number of residues.

Example of hinge motion: calmoduline

Large-scale movement of calmodulin involving splitting of one long helix. The total rotation of one domain relative to the other is upwards of 150 degrees.

Morph serverIssues Morph server steps

Get start and end configuration Searches for homolog pairs

Make them compatible Homogenizes input PDB files into a consistent format

Find flexible regions Uses homogenized structures to perform a number of additional analyses (hinge location)

Obtain motion Uses adiabatic mapping to interpolate PDB files

Visualize output Allows visualization of interpolated structures in a variety of movie formats

Quantitatively measure output Calculates statistics

Getting start and end configuration

Automated search in SCOP, PDB, etc

Submitted

Manual

Automatic

>1200 user submitted morphs

~200manually classified motions

>14000 automatically classified motions

As of 2003

Making input homogenous

• Most of the truly novel functionality of the server

• Establish equivalence of sequences: alignment (AMPS for similar proteins, structural alignment for distant proteins)

Finding hinges

• Construct a search window of 24 residues and examine each position along peptide backbone in the window. If one half in the window belongs to one domain and the other part belongs to another domain, a hinge is detected.

• If hinges are not found along the backbone, a window is reduced by two and a scan is repeated. After a scan was performed 5 times and no hinges were detected, a failure is reported.

A stand-alone application Hinge master implements several techniques to efficiently spot hinges.

find hingesaccess results

Finding hinges

Interpolation: adiabatic mapping

The morph server attempts to describe protein motions as a rigid-body rotation of a small “core” relative to a larger one, using a set of hinges. To ensure all statistics between any two motions are directly comparable, the motion is placed in a standardized coordinate system.

A pathway interpolation is produced by two principal methods:•Straight Cartesian interpolation. The difference in each atomic coordinate (between the known endpoint structures) is calculated and then divided into a number of evenly spaced steps.

•Adiabatic mapping. Energy minimization is added after each interpolation step. This procedure produces interpolated frames with much more realistic geometry.

Visual rendering

Interpolation produces a set of .pdb files that can be used to produce a custom movie

A movie is created on-the-fly:• a PDF file containing individual frames• MultiGif, Quicktime, MPEG with 2D movie• a VRML 3D movie (requires a plugin)

Statistics

• Maximum C alpha displacement, rotation angle in degrees around putative hinge regions and other parameters

• The statistics are detailed enough to perform automatic preliminary classification of the motion

New multichain morph server

• http://molmovdb.org/cgi-bin/beta.cgi• FRODA lite option added, allowing to

invoke directed dynamics with default parameters

• Hydrogen bonds are not considered, adding protons is not required

• Paths produced by FRODA avoid sterically impossible trajectories

Sample morph

Initial file and final result of FRODAsimulation from previous class were

fed to beta serverdirected motion of barnase

Alternatives

• RigiMol: • Identifies rigid or semi-rigid structural domains. • Displays RMSD, rotation, and translation statistics for the

domains. • Generates interpolated coordinates that effectively "morph"

between the structures. • Outputs standard PDB files to be used for visualization

• LSQMAN: Structural alignment program that does simple morphing using Cartesian or "internal" coordinates.

• Indie server at the NIH. Similar to the Morph Server, but uses linear interpolation and requires the submitted files to have the same sequence.

• DynDom Performs detailed structural comparisons similar to the statistics from morph server. This has a database of motions as well.

Questionable elements

• Every user is free to enter data to MolMovDB

• The system cannot accommodate all types of motions

• Quality of a morph• Quantitative measure

References

• The Database of Macromolecular Motions: new features added at the decade mark. S Flores, N Echols, D Milburn, B Hespenheide, K Keating, J Lu, S Wells, EZ Yu, M Thorpe, M Gerstein (2006) Nucleic Acids Res 34: D296-301.

• The morph server: a standardized system for analyzing and visualizing macromolecular motions in a database framework. WG Krebs, M Gerstein (2000) Nucleic Acids Res 28: 1665-75.

• Wells S., Menor S., Hespenheide B.M and Thorpe M.F. (2005) Constrained Geometric Simulation of Diffusive Motion in Proteins. Physical Biology 2, S127-S136.

• The Yale Morph Server (http://molmovdb.org)

• http://www.public.asu.edu/~frasch/molecular_simulations.htm