Tutorial
This tutorial shows how to use FilaSitus programs with VMD to greate actin filaments at low resolution in a triangular bundle and side-by-side arrangement. The created files can be compared with precomputed files provided in the FilaSitus solutions directory. More documentation is also available in the User Guide, the Methodology page of Situs, the VMD user guide, and in the published Situs articles.
Content
Creating Atomic Structures of Actin Filament Bundles with filabuild

The filabuild tool generates multiple copies of the input PDB file in the z-direction according to a user-specified helical symmetry, and in the (x,y) direction, according to a specified offset. All helical parameters, the start angle, flipping of filaments, etc., can be adjusted by the user in the configuration files.

Here we show how this is done with the provided input files. At the shell prompt, enter (you can cut and paste this): 

./filabuild nohy_start_sd1.pdb configure_36_17_bundle.dat 13_sd1_bundle.pdb
./filabuild nohy_start_sd1.pdb configure_36_17_sidebyside.dat 13_sd1_sidebyside.pdb
./filabuild nohy_start.pdb configure_36_17_bundle.dat 13_bundle.pdb
./filabuild nohy_start.pdb configure_36_17_sidebyside.dat 13_sidebyside.pdb

The first two commands create filaments from actin's subdomain 1 (SD1) only, whereas the last two generate the full structures. To learn the functionality of the program, you can inpect the resulting stuctures with VMD. You can explore the appearance of the structures depending on the specified parameters that are documented in the configuration files.

We have now created structures of full actin and SD1, both as a triangular bundle, and side-by-side. Note that the angle (phase) offsets were adjusted between bundle and side-by-side to reflect the correct orientation of the filaments when the triangular bundle is "folded" open to show the interior contacts.
Creating Low-Resolution Volumetric (Simulated EM) Data with kercon

To create a visually appealing low-resolution representation of the atomic structures, we use the kercon tool to project them to a cubic lattice and to convolve the resulting 3D density with a Gaussian blurring kernel at a user-specified resolution. Here, we take the 4 structures created above and blur them to a resolution of 20A, although smaller values can be entered to reveal more spatial detail. The choice of resolution is up to the user.

Here we show how this is done with one of the structures. At the shell prompt, enter: 

./kercon 13_bundle.pdb 13_bundle.situs

Next, at the program prompt, enter 2 (mass density blurring), the desired voxel spacing of the lattice: 6 (A), the desired kernel: 1 (Gaussian), target resolution: 20 (A), kernel amplitude: 1 (this is arbitrary, different values give a different scaling of the 3D density), sigma correction: 1. Here is the complete output of this run: 

./kercon 13_bundle.pdb 13_bundle.situs
pdbio> 124404 atoms read.
kercon> What kind of 3D density function do you want to create:
kercon>
kercon>      1: Charge density (atom charges will be read from PDB occupancy field)
kercon>      2: Mass density (atom masses are assigned automatically)
kercon> 2
kercon> There are 124404 non-hydrogen atoms, represented by 125370 equally weighted input atoms
kercon>
kercon> The input structure measures 174.508 x 163.790 x 420.297 Angstrom
kercon>
kercon> Please enter the desired voxel spacing for the output map (in Angstrom): 6
kercon>
kercon> Please select the type of kernel:
kercon>
kercon>      1: Gaussian: A exp(-1.5 r^2 / sigma^2),
kercon>         useful for resolution-lowering of atomic structures.
kercon>
kercon>      2: Hard Sphere: 0 (outside) or A (inside),
kercon>         useful for bead-modeling at reduced complexity:
kercon>         - sphere radii are read from input PDB file (B-factor field).
kercon>         - sphere boundaries are anti-aliased 1-voxel wide
kercon> 1
kercon> Kernel width. Please enter (in Angstrom):
kercon>     (as pos. value) target resolution (== 2 sigma) or
kercon>     (as neg. value) kernel half-max radius
kercon> Now enter (signed) value: 20
kercon>
kercon> The Gaussian kernel has the following properties:
kercon>
kercon>     Gaussian, A exp(-1.5 r^2 / sigma^2)
kercon>     sigma = 10.000A, r-half =  6.798A, r-cut = 17.321A
kercon>
kercon> Please enter the desired kernel amplitude A: 1
kercon>
kercon> Do you want to correct sigma for spreading introduced by tri-linear projection to lattice?
kercon>
kercon>      1: Yes (slightly lowers the kernel width to maintain target resolution)
kercon>      2: No
kercon> 1
kercon> Projecting atom masses to cubic lattice by tri-linear interpolation...
kercon> ... done. Lattice spread (rmsd):  4.244 Angstrom
kercon>
kercon> Computing Gaussian kernel (correcting sigma for lattice smoothing)...
kercon> ... done. Kernel map extent 7 x 7 x 7 voxels
kercon>
kercon> Convolving lattice with kernel...
kercon> ... done. Spatial resolution (2 sigma) of output map: 20.000A
kercon>
volio> Writing density data...
volio> Volumetric data written to file 13_bundle.situs
volio> File 13_bundle.situs - Header information:
volio> Columns, rows, and sections: x=1-41, y=1-39, z=1-82
volio> 3D coordinates of first voxel (1,1,1): (-78.000000,-78.000000,-66.000000)
volio> Voxel size in Angstrom: 6.000000

Again, the user can change some of the program parameters if desired. In the case at hand one would probably only be interested in the effect of changing resolution. As an exercise, repeat the same procedure for the input structures 13_sidebyside.pdb, 13_sd1_bundle.pdb, and 13_sd1_sidebyside.pdb, creating the corresponding output maps in Situs format: 13_sidebyside.situs, 13_sd1_bundle.situs, and 13_sd1_sidebyside.situs, respectively.
Visualizing Isocontours with VMD

Now we can load the isocontour surfaces into VMD. For convenience, the following VMD commands can be pasted directly into the VMD command console: 

mol load situs 13_bundle.situs
mol load situs 13_sd1_bundle.situs
mol modstyle 0 0 Isosurface 40 0 0 0 1 1
mol modstyle 0 1 Isosurface 35 0 0 0 1 1
mol modcolor 0 0 ColorID 0
mol modcolor 0 1 ColorID 1
display resetview

Here is the view created by this representation (note that you van make snapshots with the VMD File->Render menu):

Similarly, we can visualize the side-by-side representation (open a new VMD session): 

mol load situs 13_sidebyside.situs
mol load situs 13_sd1_sidebyside.situs
mol modstyle 0 0 Isosurface 40 0 0 0 1 1
mol modstyle 0 1 Isosurface 35 0 0 0 1 1
mol modcolor 0 0 ColorID 0
mol modcolor 0 1 ColorID 1
display resetview

Here is the view created by this representation:

We chose an isovalue of 40 because the resulting blue surfaces appear similar to the molecular surface of the actin filaments (see below). Also, we chose a slightly smaller isolevel 35 for the red SD1 surfaces so that they occlude the surfaces of the full actin filaments.
Cropping Maps with pindown

The surfaces created in the above manner are somewhat uneven at the ends of the filament, due to the effect of the blurring on the finite-size structures. One may wish to crop the density files to trim off the filament ends to create a cleaner look. This has to be done on the Situs formatted files, i.e. before the isocontours are created. The suitable map cropping tool is pindown, which allows one to extract a box of interest based on an enumeration of the lattice points. We enter one caveat here: Since the Situs formatted maps may differ in their size and origin (3D coordinates of the [1,1,1] voxel), one has to calculate at which lattice index to crop to create similar appearance in two related files (e.g. in full actin and SD1). To facilitate this calculation, the pindown program displays the map parameters such as voxel spacing, origin and the number of x, y, and z increments.

For a clean cropping of the ends on the data at hand we suggest the following: At the UNIX shell prompt enter e.g. (similar for the other files): 

./pindown 13_bundle.situs 13_bundle_crop.situs

At the pindown prompt enter the following crop regions: 1-41,1-39,15-67 (13_bundle_crop.situs), 1-41,1-39,13-65 (13_sd1_bundle_crop.situs), 1-61,1-27,15-67 (13_sidebyside_crop.situs), and 1-61,1-27,13-65 (13_sd1_sidebyside_crop.situs). One can then visualize the cropped density data (detail view):



For more general map editing purposes we recommend the newer tools (such as voledit, volhist, voldiff, etc) distributed with Situs.
Creative Use of the VMD Clipping Plane

The side-by-side representation is useful as it allows one to clip away any unwanted parts of the molecular surfaces with the VMD clipping plane. To test this, load again 13_sidebyside.situs  and 13_sd1_sidebyside.situs as above. Now use the VMD "Far Clip" plane (VMD Display Menu / Display Settings) and bring it closer to the front. Take a snapshot with the VMD Render tool. Next toggle off 3_sd1_sidebyside.situs and set the clipping plane behind the scene. Take a second snapshot. In image editing programs like Adobe Photoshop you can make the black background of the first snapshot transparent and superimpose the colored pixels as a layer onto the second snapshot. This should look as follows:

+ =

This is a quick way to viualize the SD1 contacts in the center of the triangular bundle, if the phase offset angles of the side-by-side representation were chosen accordingly.

Return to the front page.