Fleming Lab
Computing
Our lab is an experimental lab but we frequently use computational methods to help analyze and interpret the experimental data. Below are some of the computing applications we have developed to calculate various properties of biological macromolecules.
HullRad
HullRad is a computer program and webserver that calculates hydrodynamic properties of macromolecules from structure files.
It uses a convex hull representation of the molecule. The figure at left shows a bacterial 50s ribosomal subunit enclosed by the initial convex hull used by HullRad.
REDUX
Torsion angle Monte Carlo protein molecular simulations with reduced residue representation.
The figure at left shows:
(TOP) A standard VDW sphere representation of an oligopeptide with all twenty amino acids.
(MIDDLE) A REDUX model representation with a single pseudoatom side chain, a Cα atom, and a single pseudoatom peptide bond for each residue. The size of the side chain pseudoatom is proportional to the volume of the respective side chain.
(BOTTOM) Both models superimposed with the REDUX model as semi-transparent surface.
REDUX is useful for converting an extended polypeptide (e.g. as built in PyMOL) to a collapsed conformation of the protein (e.g. image at left) suitable for further simulation in a standard molecular mechanics force field. It's a facile method to make unfolded, but collapsed protein models.
REDUX is a fast Python script and the final protein conformation can be controlled by adjusting the following "force field" terms:
- Soft debump (V = Eps,i,j[(Sigma,i,j/ri,j)**12]
- Residue specific pairwise contact
- Radius of gyration (Rg) confinement
- Backbone hydrogen bond strength
- Backbone dihedral distribution (directed Ramachandran plot distributions)
Download the distribution as a tarred and gzipped archive file.
A userguide in PDF format is available.
Occluded Surface (OS)
Occluded surface is defined as the atomic surface that is less than 2.8 Å from the surface of neighboring non-bonded atoms. That is, if a water molecule cannot fit between two atoms they occlude each other. Occluded surface is similar to buried surface but is more sensitive to packing geometry than buried surface identified using a rolling probe.
Occluded surface patches of each atom may be depicted by extended surface normals as in the image at left. A sense of the atomic packing complementarity can be gleaned from such a depiction. Output includes the number and lengths of occluded surface normals for calculation of relative packing complementarities. This type of analysis is also useful for investigating protein-protein interactions.
Alternately, occluded surface patches of each atom may be depicted as dots as in the image at left.
Summation of the dot surface areas provides an alternative to buried surface area calculated by rolling probe methods.
Download the distribution as a tarred and gzipped archive file.
The distribution contains more details and instructions in HTML format.
A new version of OS, called FIBOS, is available through a collaboration with the da Silveira lab in Brazil.
The original OS used equatorial slices to allocate the surface normals (left,top), a method that introduces some inhomogeneities
in the surface areas associated with the normals. FIBOS uses Fibonacci spirals to allocate the
surface normals (left, bottom). This new method provides randomly distributed surface normals with a narrow distribution of associated surface areas.
and
https://github.com/insilico-unifei/fibos-py.
Ref: 1
Difference Distance Matrix Plot (DDMP)
When one has two high resolution macromolecular structures that are slightly different in conformation (e.g. different ligand binding, or different crystallization conditions) it is advantageous to be able to identify those regions of the structure that are different in the two structures. This is where DDMP can be useful.
The figure at left shows a typical plot from DDMP. The difference distance matrix of a ribonuclease mutant (1rbe) was subtracted from the difference distance matrix of ribonuclease (1rnv). The delta values are plotted with colors representing Cα differences of -1.25 to 1.25 Å according to the scale at the top.
The red residue numbers represent those areas of the protein that are more expanded in 1rbe compared to 1rnv.
Download the distribution as a tarred and gzipped archive file.
The distribution contains more details and instructions in HTML format.
