I am using
pymolto visualise the secondary structure of protein using its
pdbfile comes from a simulation, which contains multiple frames. After loading the
pdb, its secondary structure (e.g. sheet, helix) could not been recognised. Surprisingly, if only one frame is kept, its secondary structure could be seen. So how to enable secondary structure recognition for a
pdbfile with multiple frames?
An example of the
pdbfile with multiple frames is shown below.
REMARK GENERATED BY TRJCONV TITLE Protein in water t= 0.00000 REMARK THIS IS A SIMULATION BOX CRYST1 116.453 116.453 116.453 90.00 90.00 90.00 P 1 1 MODEL 1 ATOM 1 N ASP 1 85.582 59.777 48.367 1.000.0000 N ATOM 2 H1 ASP 1 84.882 59.067 48.507 1.000.0000 H ATOM 3 H2 ASP 1 85.162 60.617 48.747 1.000.0000 H… ATOM 6615 OT ALA 442 28.032 36.877 69.157 1.000.0000 O ATOM 6616 O ALA 442 30.092 36.087 68.677 1.000.0000 O ATOM 6617 HO ALA 442 30.072 35.867 69.597 1.000.0000 H TER ENDMDL REMARK GENERATED BY TRJCONV TITLE Protein in water t= 1000.00000 REMARK THIS IS A SIMULATION BOX CRYST1 116.384 116.384 116.384 90.00 90.00 90.00 P 1 1 MODEL 2 ATOM 1 N ASP 1 75.052 41.097 56.132 1.000.0000 N ATOM 2 H1 ASP 1 75.622 41.407 55.352 1.000.0000 H ATOM 3 H2 ASP 1 75.602 41.357 56.932 1.000.0000 H ATOM 4 H3 ASP 1 74.682 40.167 56.272 1.000.0000 H ATOM 5 CA ASP 1 74.032 42.157 56.202 1.000.0000 C… ATOM 6615 OT ALA 442 45.292 49.247 90.922 1.000.0000 O ATOM 6616 O ALA 442 47.102 49.617 89.632 1.000.0000 O ATOM 6617 HO ALA 442 47.662 49.327 90.342 1.000.0000 H TER ENDMDL
It seems that if no
HELIXrecord is present in the PDB, PyMol attempts to assign secondary structure itself. For whatever reason, it does not do this if multiple
MODELs are present. You should be able to use the
dss command to force PyMol to calculate secondary structure in a PDB with multiple
Since I am obsessed with actin and actin-binding proteins, I will use the complex of actin with profilin for this demonstration. Profilin acts as an ATP exchange factor for G-actin. “Old” actin monomers from recently depolymerized actin filaments have ADP bound to them, and upon binding to profilin, will exchange the ADP for ATP. I found a nice structure (determined by X-ray diffraction) of actin bound to profilin on the RSCB Protein Data Bank website. Open PyMol and type the following into the Pymol command line:
Extract the various part of the complex that we want to work with.
Get the interface residues script and identify the interacting residues
Go to the InterfaceResidues script website and scroll down the page until you reach the box The Code . Copy the code and paste it into a new text file that you name InterfaceResidues.py . Be sure to use a plain text editor and save the files as plain text so that no hidden or extraneous characters are included in the file. My favorite free text editor is TextWrangler. Even though you are saving the file as plain text, make sure that the file extension is .py and not .txt . Save the script to your PyMOL working directory.
The InterfaceResidues script relies on the interacting proteins have chains labeled A and B . Check that the complex has the correct chain labels.
Because the chains are not named A and B, I have to create a complex with the 2 chains to use the InterfaceResidues script. Run the script, create the complex, and identify the interacting residues.
The script will create a selection named (interface) that contains the interface residues.
More visualization with VMD & PyMOL
As a continuation of the previous tutorial, I will show you a few more things you can do with VMD along with some tricks for PyMOL. We’ll use a model for the phosphopantetheine tether of acyl carrier proteins. This prosthetic tether forms a thioester linkage to amino acids and delivers them to the active sites of many different enzymes found most commonly in natural product biosynthetic pathways.
Nevertheless, all you really need to know is that it is the example molecule we’ll work with today, and you can download the coordinates we’ll use for the visualization here and here.
1.Revisiting VMD: Trajectories, Tachyon and beyond.
Now, there are a few details of VMD we didn’t get to last time that are useful for getting more descriptive figures from your coordinate files with VMD. Start by opening the PPant-Anim.xyz file in VMD. When you open this file, you’ll notice that there is more than one set of coordinates in it. You can look at the different coordinates by sliding back and forth the bar on the bottom of the VMD main window. You’ll notice that the different structures are not aligned in any way.
Here’s how to align structures in VMD.
Go to Extensions>Analysis>RMSD Trajectory Tool and you’ll see a window that looks like this:
elements, the backbone of a protein, and use expressions that exclude certain elements or indices. Now that we’ve aligned the structures, when we use the bar on the main VMD window, we can see that the structures are all the same at the phospho end of the molecule.
How to visualize all of the structures at the same time:
In VMD, we can also overlay all of the different structures from our trajectory at the same time. Go to Graphics>Representations and select the Trajectory tab. Since there are 15 frames numbered from 0 to 14 and we want to see all of them, we’ll write 0:14 in the Draw Multiple Frames dialog box.
Now we can go back to the Draw style tab and choose coloring method as Timestep. The color gradient by default is red/white/blue. You can choose an alternative option by going to Graphics>Colors and choosing the Color Scale tab (I’ll actually be using red/green/blue today with an offset of 0.30 to make the colors lighter). As for the rest of the settings, see the previous tutorial for turning on ambient occlusion, setting the material to diffuse, and Drawing Method as CPK with Sphere Scale/Bond Scale 1.6 and 1.0.
Now, the result of our rendering can differ greatly depending on which renderer we use. To give a soft appearance that emphasizes depth, we can use Tachyon with a diffuse material. This nearly replicates Qutemol-like rendering, but you’ll want to play with the colors that you use. In order to show the differences between the two rendering engines more clearly, look at the following scene rendered with Tachyon (left) and POV-Ray (right):
Note, another useful technique in visualizing with VMD is to assign different materials to different elements or selections. We can change the selected atoms in representations for our CPK rep to “ All not element H ”. Then we create a new VDW rep with sphere scale 0.4 and selected atoms are chosen as “element H”. For materials, we select Glass3 and set the coloring to Timestep and turn on visualizing all frames again in the Trajectory tab. The end result, rendered with Tachyon is (click on image for a larger view):
2. PyMOL: Great illustrations and not just for proteins.
PyMOL is a molecular visualization tool that is based upon a Python base and works with Python scripting . The wiki for PyMOL is a great resource that delves into many details that I’ll be glossing over today. For PyMOL , open the PPant.xyz coordinates by going to File>Open and finding the file.
Note: PyMOL supports multiple frames in a single file, but only if they are in PDB model format. You can achieve this on PPant-Anim.xyz by converting the file with openbabel , if you like.
We now want to change the color of our backbone atoms. In PyMOL, each model, as it is opened up, has a different default backbone color that is used to distinguish it from the previous model. In order to change that color, select the small C and choose the color. I chose density under the blues sub-menu. Then click C again and choose by element with the top selection CHNOS with C greyed out. This will return the colors of the other elements to typical defaults.
When it comes time to render, we can choose ray trace modes from the commandline of our PyMOL window demarcated as PyMOL> and type:
set ray_trace_mode, <# = 0,1,2,3>
where 0 is the default, 1 is a cartoon-like flattened representation, 2 is black and white cartoon, and 3 is 1 but more extreme. We’ll choose 3 today.
Now, to render our final image, we type “ ray ” at the PyMOL> commandline. Once the rendering finishes (which you can tell from the progress bar), you’ll see the final image.
Click File>Save Image As. >PNG to open a dialog box and save your final image.
I hope that this tutorial has helped you with using VMD and PyMOL and rendering with POV-Ray and Tachyon. Please email me if you have any additional questions not answered here!
Working with PyMOL
PyMOL and VMD are the tools to go (at least for me) if it comes to visualizing proteins and their cofactors, from a PDB-file (What a PDB file is you can learn here.). This is why I’m going to devote this post to PYMOL and one of the next to VMD.
First of all, if you want to play with it as well go to: https://pymol.org/2/
Note1: There is a whole wikipedia on PyMOL and its commands, so please check this out: PyMOL-Wiki Settings
Note2: “foobar” is often used as a spaceholder for a name.
Really impressive is also the Gallery of the PyMOLwiki page, with many recipes to cook for yourself.
As many shell based programs, also pymol has a runtime configuration (short rc) file, where certain predefined parameters can be tweaked.
Many pymol functions listed here can be easily realized (maybe even more easy) by using the build in GUI features, that can be found above or at the sides of the program. But pymol is also a console program, which allow us to use specialized scripts to execute more commands at once. Pymol specific scripts have the extension .pml and allow for the collection of commands. When loading them in, there will be executed first.
But now let get into it and collect some commands.
To load in a new pdb code simply use
Selection & Extracting
Selection allows you uniquely label a set of atoms that are part of an object (e.g. a protein that you loaded in).
For example, select all the C-alpha-atoms and name it foobar
Extraction on the other hand, creates a new object out of the selected set of atoms.
For example, you may want to extract the ligand (in this case NAD) of a protein extract to manipulate it’s position later on.
PyMOL employs specific identifiers (and useful abbreviations) for selecting specific things.
- chain : (c.) : chain specific
- pepseq : (ps.) : residue by name (1-letter-code)
- resn : (r.) : residue by name (3-letter-code)
- resi : (i.) : residue by index
- name : (n.) : atom by name
- index : (idx.) : atom by PyMOL internal index
Select all the C-alpha, C, N and O-atoms and name it foobar
PyMOL additionally employs a couple of pretty strong selector word (and their short forms are even better).
- all : (*) : all atoms
- hydro : (h.) : all hydrogen atoms
- hetatm : (het) : all HETATM records
- visible : (v.) : all enabled and visible atoms
- present : (pr.) : all atoms with defined coordinates
As an example, this code selects all HETATM records, which includes all cofactors, waters, ligands.
Select residue specific atoms, e.g. C-beta and C-gamma of residue 4.
Select all heteroatoms in the molecule that are not waters.
Selects all residues within 5 Angstrom around residue 25 (which can be an ATOM or a HETATOM or whatever) and name it foobar
Another way to do it, would be:
Logical operations are also possible in the selection context.
- not S1 : !S1 : NOT
- S1 and S2 : S1 & S2 : AND
- S1 or S2 : S1 | S2 : explicit OR
- S1 or S2 : S1 S2 : implicit OR
- S1 & (S2 S3) : Parenthesis - Order
- first S1 : first atom
- last S1 : last atom
Select all alanine in the large subunit (L) of the protein.
As your selection macros are getting larger with more subunits, residues and atoms, you can use Selection Macros, which have the following format:
Selection Macros have to meet a certain format:
- must contain at least one slash
- no spaces allowed
- empty fields are interpreted as wildcards, so since many PDB-files don’t have a segment identifier, the “seg” part can be left alone
A real simple way to see everything about your protein of interest (especially with their ligands) is by using the GUI:
Action on your protein of interest –> preset –> ligand sites –> cartoon
The most popular representation of proteins three-dimensional structure, the cartoon, can be tuned quite nicely (PyMOL Wiki).
Here you can set the length and width of the alpha-helices:
And the same goes for the beta-sheets:
And of course for the loops:
But there are also many different cartoon types available.
For “type” you can use many different things:
automatic , loop, rectangle, oval, tube, arrow, dumbbell, putty, das
My favorite representation is the dumbbell, which includes the fancy_helices option. Once you execute this command, you are in the dumbbell mode:
Here you can also set the length and width of the helices, but via a different command:
Additionally you can set the radius of the dumbbell (at the edges of the alpha helices)
Force PyMOL for strictly respect secondary structural elements color-wise:
Make the cartoon transparent (values between 0 and 1) :
You can also color certain regions differently than others:
You can also represent the protein as spheres.
Set the sphere size to 0.25 A.
The scientifically most accurate representation, especially if you want to look at the single atoms possibly interacting with another, then you should use the “sticks representation”.
Set the stick radius to a value between 0 (invisible) and 1 (full).
Change the transparency of the sticks between 0 (opaque) to 1 (invisible).
Another really nice representation especially for visualizing protein pockets, is the surface representation. Here it might be beneficial of creating two objects (the protein and the ligand) using the extract method:
Before you should make sure that your ligand is what should be “carved around”:
And of course there are many command to make your surface look pretty, like changing the color.
Or the surface type from dots (1), to a mesh (2) or a complete surface (3).
And of course there are many command to make your surface look pretty.
Labeling of structures
Again there is a really good Wiki Page available on this topic.
After your selection have been made and your crystal structures look as you want them to look like, you can go ahead and label certain residues with their molecule name, atom name, charge or b-factor. For example
Here also you can use the Selection Macros, to give certain atoms specific names.
For me the labels are often to small, therefore I change it to 20:
If you don’t like the position of your label, you can change it:
If you want to see your double bonds, use:
If you want to see your ligands in your protein well lit.
Count all atoms of a selection called “foobar”.
Create a pseudoatom in the center of a selection.
Set a specific color named foobar using the Normalized RGB definition.
Color a selection (protein_foobar) with your defined color (color_foobar).
This can also be applied specific for certain secondary structures (ss).
Pymol can be used in the stereo mode to view your protein in 3D.
For aligning two structures based on sequence similarity, use the align command:
On the other hand, if you’re planning on simply superimposing two structures independent of sequence similarity, use the super command:
Mutagenesis & Altering of the PDB file.
Mutagenesis, the changing of amino acids in the PDB file can be best doing by using the Mutagenesis Wizard (Wizard –>Mutagenesis –> Proteins).
The way to mutate one or more amino acids is pretty straight forward and outlined here.
Unbonding a chemical bond can be useful sometimes, since some PDB-files are not well resolved and some mistakes might have happened. Here it is much easier to use the integrated Builder-tool (that you can find the right top corner of the program). But there is also a command for that. In this case the N10 nitrogen of FAD was accidentelly bonded to a hydrogen of the O7 oxygen of an NADP. Here the Selection Macros (also mentioned above) are the way to go.
If you want to repeat the same operation over and over again, you may want to use the Iterate function.
Change the name of chain A to chain C.
After having deleted parts of your protein (e.g. the first 100 AS) you might want to change the numbering of the sequences again, to start at 1 again.
After having altered your PDB file (the numbering, chain identifies…), use sort to get a newly ordered view on the sequence.
You can also change the coordinates of parts of your protein.
After having altered your PDB coordinates, use rebuild to implement the changes (sometimes it is implemented in real time without the need of rebuild).
How to make pretty images in PyMOL I’ve covered in seperate post, which you can find here.
Catalases are tetrameric heminic enzymes that decompose hydrogen peroxide into water and oxygen. They serve to protect cells from the toxic effects of hydrogen peroxide. As a case study, we used ENDscript to decipher features in the crystal structure of Proteus mirabilis catalase (PDB entry 2CAH) ( 21).
The two fundamental residues of the heme-iron catalytic site are a distal histidine and a proximal tyrosine. A unique methionine sulfone is observed in the distal site of P. mirabilis catalase. Another peculiarity of this bacterial catalase is its ability to bind nicotinamide adenine dinucleotide phosphate (NADPH) for the prevention of inactivation by hydrogen peroxide ( 22).
The first flat figure clearly shows that the catalase structure has an α+β topology (Figure 1A and Supplementary Data 2). Moreover, the N-terminal region is involved in extensive protein:protein crystallographic contacts as shown by italic A letters below the sequence block. A red letter identifies a contact < 3.2 Å while a black letter identifies a contact between 3.2 and 5 Å. Leu31 is positioned along a 2-fold crystallographic axis as shown by an italic hash symbol. Leu40 and Asp44 are in contact with the heme group of a symmetric monomer as shown by an italic colon. Asp44 is also involved in a protein:protein crystallographic contact as shown by a blue frame. Arg51 binds the heme group of the monomer as shown by a normal colon on a light yellow background. The non-standard residue in position 53 (labeled X) is the methionine sulfone of the distal site. Phe140 may be a critical residue: it has close contacts with the heme group as shown by the red colon. It is also involved in protein:protein contacts as shown by the blue frame. Finally, it is involved both in crystallographic and non-crystallographic contacts as shown by the orange background. His173 binds tightly NADPH as shown by a red caret. The second flat figure (Figure 1B and Supplementary Data 3) reveals that secondary structure elements are well conserved in catalases deposited in the PDB.
The 3D ‘Sausage’ representation (Figure 1C and Supplementary File 1) allows for emphasizing the Cα trace of the C-terminal region that varies between catalases (the tube radius is proportional to the mean rms deviation between 2CAH and homologous catalases). In agreement, the C-terminal region is poorly conserved in sequence and is mainly colored in white (color ramping from white, low conservation, to red, identity). By contrast the N-terminal region, which protrudes out of the protein core, is surprisingly well conserved. In fact, this region is deeply buried in the biological tetramer, this being visible by displaying the biological assembly (Figure 1D and Supplementary File 1) automatically incorporated by ENDscript in the PyMOL session file. Finally, thanks to a preset button on its control panel, PyMOL can display the protein solvent accessible surface colored according to the level of sequence conservation (Figure 1D).
I’m very happy to learn that Zeynep Kürkçüoğlu Soner, a scientist working on generating atomistic conformers for proteins with various size, saw this tutorial and adapted it for her purposes on PyMOL. Here is her method with her own words:
It is generally hard to visualize protein motions on static images, contrary to movies playing sequentially the snaphots obtained from simulations such as molecular dynamics. For my thesis and the paper that is related with it, I was trying to find a means to clearly present such motions, and as Cem mentions in his tutorial, the vectors are not always the best way. Then, I encountered Cem’s work related with “visualizing protein motions with static images” and realized that the representation that I have been looking for was right in front of me! In my work, I do not have a specific direction vector as in the case of normal vectors, however I thought that with a little modification, the method can be suitable for my figures as well. Voilà, the result for p38 kinase:
The figure is obtained from the software PyMOL, as many structural and computational biologists are familiar with (like VMD). Personally, I like the colors and shapes obtained from PyMOL more than the ones from VMD.
Analysis of data from molecular dynamics simulation
When the simulation has finished, it is time to continue with the analysis of the data. The analysis comprises of three stages. First, it is necessary to perform some standard checks to assess the quality of the simulations. If the results from these analyses show that the simulations are fine, then the analyses required to answer the research question asked can be performed per simulation. Finally, the results from different simulations can be combined.
NOTE: The names of the files should be self-explanatory, depending on the system you simulated. Here we assume that the default names have been used, which means that there are the following files:
- topol.tpr : The run input file containing a complete description of the system at the start of the simulation
- confout.gro : Structure file, contains the coordinates and velocities of the last step
- traj.trr : The full precision trajectory containing the positions, velocities and forces over time
- traj.xtc : A light weight trajectory, containing only coordinates in low precision (0.001 nm)
- ener.edr : Energy related parameters over time
- md.log : A file containing information about the simulation.
As a side note, many of the analysis tools produce graphs in the form of .xvg files. These files can be viewed with the program xmgr or xmgrace, but they can also be viewed in a terminal using the python script xvg2ascii.py.
Comparison to experimental structural data
We can use a PCA not only to analyse an MD simulation and extract the main concerted motions from a trajectory, but also for a comparison of multiple simulations or for the comparison to an ensemble of experimental structures, in terms of dominant, collective motions. For T4 lysozyme, a large number of x-ray structures has been determined (if you search the PDB for "T4 lysozyme", you'll find more than 400 structures). We have collected 38 non-redundant x-ray structures of T4 lysozyme, which are all together collected in allpdb_bb.xtc. Just like on the simulated trajectory, we can perform a PCA on the collection of experimental structures:
Question: What kind of motion is represented by the first and second eigenvector? How does this compare to the results from the simulation?
For a quantitative comparison of the simulation and the experimental strutures, we'll project both the experimental structures and the simulated structures onto the eigenvectors extracted from the x-ray structures:
Question: How do the x-ray and MD ensemble compare to each other? Why do you think the MD simulation does not sample the x-ray configurations on the extreme left of the plot (the most open x-ray conformations)? And why do you think the twisting mode is sampled with a larger amplitude in the simulation than in the ensemble of x-ray structures?
Two additional simulations are available for T4 lysozyme, that started from different x-ray structures. Download the two backbone trajectories to your local disk: md2_backbone.xtc and md3_backbone.xtc. Cross project also these two trajectories onto the x-ray eigenvectors and visualise all three simulations and the ensemble of x-ray structures projected onto the first two eigenvectors of the x-ray ensemble.
Question: Which of the three simulations covers most of the crystallographic structures? Do you consider the simulations long enough to sample the most relevant conformations of the protein?
Question: All three simulations were of the substrate-free protein. Would you expect a more 'open' or a more 'closed' state as the most probable configuration in this situation?
Question: Alternatively, we could have carried out a Normal Modes analysis of this protein, approximating the energy landscape by a multidimensional harmonic energy minimum. Does the motion along the first PCA eigenvectors appear (quasi)-harmonic? If yes, which force-constant would you estimate?
Can pymol show cartoon (secondary structure) for a pdb of multiple frames? - Biology
UCSF ChimeraX is the next-generation molecular visualization program from the UCSF RBVI. Images in this page are CC0 and can be reused freely, although we encourage citing UCSF ChimeraX by name.
Coulombic Electrostatic Potential
Coulombic electrostatic potential (ESP) can be calculated and displayed with surface coloring using the command coulombic or the Molecule Display icon . No separate calculation or input ESP file is required. The image shows the first assembly defined for PDB 3eeb, the protease domain of a toxin from Vibrio cholerae, with the default Coulombic coloring: red-white-blue over the value range &ndash10 to 10. For image setup other than orientation, see the command file coulombic.cxc.
Assemblies from mmCIF Symmetry Information
Multimeric assemblies defined in a structure's mmCIF file are automatically listed when the file is opened and can be reconstituted with the sym command. The complex of selenocysteine synthase and tRNASec (PDB 3w1k) is shown as molecular surfaces of different colors for different chains. The asymmetric unit is on the left and the assembly specified in the mmCIF file is on the right, with arrow and text annotations from 2dlabels. See the command file sym.cxc.
Alternatively, biological assemblies can be fetched by specifying rcsb_bio or pdbe_bio as the source database.
The Matchmaker tool (or matchmaker command) is convenient for superimposing related structures without having to worry about numbering or missing residues. It superimposes proteins or nucleic acids by creating a pairwise sequence alignment, then matching the sequence-aligned residues in 3D. Secondary structure helps guide the sequence alignment for better performance on more distantly related proteins with harder-to-align sequences. By default, the fit is iterated to exclude structurally dissimilar regions and superimpose the most similar parts more closely.
The resulting sequence alignments can be displayed, as in this example of three pectate lyases: PDB 1jta, 1bn8, and 2pec (see the command file peclyases.cxc). In the sequence alignments, residues used in the final fit iteration are enclosed in light orange boxes. RMSD values and other fit statistics are reported in the Log.
Coloring by Sequence Conservation
Atomic structures, including cartoons and molecular surfaces, can be colored by the conservation in an associated multiple sequence alignment. The figure shows a structure of the &beta2-adrenergic receptor signaling complex (PDB 3sn6) with receptor cartoon colored blue&rarrwhite&rarrred from least conserved to most conserved. The &beta2-adrenergic receptor is a member of the class A G-protein-coupled receptor superfamily. Conservation was calculated from a superfamily alignment from PASS2 using the entropy-based measure from AL2CO (included with ChimeraX courtesy of Pei and Grishin). The sequence alignment and step-by-step instructions for making this image are given in the Coloring by Sequence Conservation tutorial.
Front/back (rotatable) clipping can be applied selectively to some models but not others. This is most often used to slice a molecular surface but not the corresponding atomic structure.
For example, the protein in PDB entry 1g74 has an oleic acid residue OLA in an interior pocket. The script in pmc.cxc shows the protein surface, activates front clipping for all models, and then turns it off for just the atomic model, as shown in the figure. The clipping plane can be translated and rotated interactively with the mouse .
Multichannel Light Microscopy
3D images and time series from multichannel optical microscopy are shown in the Volume Viewer tool, with easy access to hiding/showing individual channels, changing their colors, and adjusting threshold levels with the mouse. The menu of style options includes &ldquovolume&rdquo (translucent blobs, as in the image), surface, mesh, maximum intensity projection, single plane, and orthoplanes. For convenience, the step size, region bounds, and display style of different channels of the same dataset are coupled, in that changing the setting of one channel automatically changes it for the others.
The image shows human induced pluripotent stem cells, with plasma membrane in violet red, EGFP-tagged fibrillarin (as a marker for nucleolus) in yellow, and DNA (nucleus) in turquoise. The data are publicly available from the Allen Cell Explorer website, dataset: AICS-14_0.
ChimeraX virtual reality works with HTC Vive, Oculus Rift, and Samsung Odyssey systems (those supported by SteamVR). Any structures, maps, etc. that can be displayed in ChimeraX can be viewed in the headset and manipulated with the hand controllers. Icon toolbars visible in the headset allow changing the scene display or hand-controller button assignments with a single click. Besides rotation, translation, and zooming, useful functions include labeling, distance measurement, bond rotation, placing markers into a map, and changing map contour levels. Virtual-reality (VR) mode can be turned on and off with the vr command, and the meeting command allows multiple users to share a single session in VR.
Curved Helix Cylinders
Protein &alpha-helices can be shown as curved-cylinder &ldquotubes&ldquo with the cartoon style command. Helix tube mode is an alternative to the standard spiraling ribbons, and both modes are fully integrated with coil and &beta-strand cartoons. The structure at left is an AMPA-subtype glutamate receptor bound to the antiepileptic drug perampanel (PDB 5l1f). The receptor is tetrameric, and each chain is rainbow color-coded from blue at the N-terminus to red at the C-terminus. Four molecules of perampanel (pink) are bound near the bottom, between the transmembrane domain and the rest of the receptor. For image setup other than orientation, see the command file ampar.cxc.
Interactive Chain Network
Chain-chain interfaces can be identified by buried surface areas and displayed as a network diagram with the interfaces command or the Molecule Display icon . In the diagram, nodes (circles) represent chains, larger for greater surface areas, and edges (lines) between nodes represent chain-chain interfaces (default &ge 300 Å 2 buried area). Dotted lines represent interfaces smaller than half the size of the largest in the structure. Diagram context menus enable a variety of actions, such as &ldquoexploding&rdquo the structure by moving chains apart, hiding all but the chains in contact with a given chain, and showing a more detailed plot of the residues forming a given interface.
The structure is an HIV envelope glycoprotein trimer bound by three copies of a broadly neutralizing antibody (PDB 5v8m), with chain information shown in the Log. Glycosylations (not displayed) were included in the surface area calculations. For setup, see the command file trimer-network.cxc.
Struts for 3D Printing
Structures can be reinforced for 3D printing with pseudobonds. In this DNA-transcription factor complex (PDB 5ego), proteins are shown as ribbons and the DNA as a molecular surface. The pseudobonds in blue were read in from a manually created file, 5ego.pb, and further reinforcements in light gray were added automatically with the struts command. For image setup other than orientation, see the command file struts.cxc.
Rotamers and Swapaa Virtual Mutation
Rotamers is an interface for showing amino acid sidechain rotamers and optionally replacing the original sidechain, also implemented as the swapaa command. The rotamers can be shown all at once, as in the figure, or individually by choosing rows in the dialog.
The figure shows binding-site residues of the thyroid hormone receptor &beta with hormone bound, PDB 3gws. Rotamers for the hormone-resistance mutations N331H and L346R are shown as partially transparent sticks, with H-bonds (light blue dashed line) and clashes (light purple dashed lines) calculated for the histidine rotamers at position 331. The rotamer-list dialog for this position is also shown. Command script rotamers.cxc contains the initial, noninteractive part of the setup.
These mutations are described in Cardoso et al., Endocrine (2020). Although one histidine rotamer may be able to form the same pocket-stabilizing H-bond as the wild-type asparagine, it also clashes with several atoms (third row in the dialog). H-bonds and clashes are not shown for the arginine rotamers at 346, but they all clash significantly with the hormone and/or other pocket atoms.
A &ldquopalette&rdquo or ordered series of colors is used to color items sequentially (rainbow) or by values such as density. The ten chains in PDB 5o3l (paired tau filament) have been colored with the commands shown as 2D labels in the images. The first two examples at left use predefined palettes (credit to www.ColorBrewer.org, color specifications and designs by Cynthia A. Brewer, Pennsylvania State University), whereas the third shows specifying colors individually.
Multichain Comparative Modeling
Modeller Comparative is an interface to Modeller for comparative (&ldquohomology&rdquo) modeling of proteins and protein complexes.
The example shows modeling the human (shades of blue) from the mouse (brown and tan) complex of programmed death-1 (PD-1) with its ligand PD-L2, PDB 3bp5.
Comparative modeling requires a template structure and a target-template sequence alignment for each unique chain. The sequences of human PD-1 and PD-L2 targets were fetched from UniProt and associated with the corresponding chains in the template structure, see model-pdl-setup.cxc. (Pairwise or multiple sequence alignments could have been used, but in this case, the template structure was simply associated with the target sequence.) Sequence-structure association shows mismatches in the Sequence Viewer: pink boxes for sequence differences between mouse and human, and gray outlines around the parts missing from the structure.
Three models were made with with default settings (other than the number of models), and the best-scoring model is shown. Two positions where sequence differences change the interfacial H-bonds are displayed.
- inactive structure (PDB 3c4f, chain A)
- activated structure (PDB 3gqi, chain A) with phosphorylated tyrosines and bound ATP analog
Morphing and other setup was done with the command file kmorph-prep.cxc, followed by interactively positioning the structure and saving the view with the command view name p1 (generally a session would also be saved at this point), then running kmorph-play.cxc to add 2D labels and record the movie.
Coloring by Molecular Lipophilicity Potential
Molecular lipophilicity potential (MLP) can be calculated for a protein and displayed with surface coloring using the command mlp or the Molecule Display icon . The image shows the photosynthetic reaction center from a purple sulfur bacterium, with MLP coloring on the molecular surface and membrane boundaries from OPM (Orientations of Proteins in Membranes entry 1eys). Blue and red balls represent the cytoplasmic and periplasmic sides of the bacterial inner membrane, respectively. Parts of the L, M, and H chains span the membrane, whereas the cytochrome subunit sits on the periplasmic side, at the top. The surface coloring ranges from dark goldenrod for the most hydrophobic potentials, through white, to dark cyan for the most hydrophilic. Ligands including lipid, detergent, heme, and various other cofactors are shown as purple surfaces.
For image setup after the structure from OPM has been opened, see the command file mlp.cxc.
Interactive H-Bond Histogram
Hydrogen bonds (H-bonds) can be identified with the command hbonds or the Molecule Display icon and plotted as an interactive histogram with the command crosslinks histogram.
The ChimeraX graphics window shows the complex between a natural killer cell receptor 2B4 and its ligand CD48 (PDB 2ptt). The receptor protein is blue, the ligand protein pink, and H-bonds between them dashed yellow, with H-bonding residues labeled. Although not done here, the H-bonds could also be labeled by distance.
The histogram of H-bond distances on the top right is interactive: when the cursor is placed over a bar in the histogram, the corresponding H-bonds are temporarily enlarged in the 3D view and the others hidden. For image setup other than orientation, see the command file hb3.cxc.
Fitting to Density Maps
Atomic structures and/or maps can be fit rigidly into other maps with the Map icon , the Fit in Map tool, or the fitmap command. A common use is to locally optimize the fit after initial placement by hand. The command includes options for symmetrical and sequential fitting, as well as global search to generate multiple starting points for local optimization.
Elongator is a highly conserved complex that associates with RNA polymerase II during transcriptional elongation. In the image, one of its six subunits, Elp2 (PDB 5m2n), has been fit into a map of the Elp123 subcomplex (EMDB 4151).
Ambient Occlusion Molecular Surfaces
Several lighting modes are available, including ambient occlusion. The image shows hemoglobin (PDB 4hhb) with the four chains shown as surfaces of different colors and heme residues as spheres. The command lighting soft or the Graphics icon can be used to turn on ambient shadowing from 64 directions. The command lighting gentle gives a similar result, except tuned to emphasize larger indentations.
The Presets menu includes a few different combinations of cartoon and nucleotide styles, shown here along with overall settings from the Publication preset plus lighting depthcue false. The style settings of cartoons and nucleotides can be controlled individually (and with many more possibilities than shown here) with cartoon style and nucleotides, respectively. See also: Toolbar nucleotides icons
ribbons/slabs cylinders/stubs licorice/ovals
Different representations of nucleotides can be shown with the nucleotides command or Toolbar icons. Options include filled rings, slabs for bases (box, muffler, or ellipsoid shape), bumps on slabs to show base orientation, simple tubes instead of ribose atoms, and continuous or broken ladder rungs. Nucleotide representations can be the same color as the ribbon or a different color, and multiple nucleotide styles can be used within a single structure.
CryoEM Ambient Occlusion
A cryoelectron microscopy map of the 26S proteasome (EMD-4321) is shown at the author-recommended contour level in two different lighting modes: &ldquosimple&rdquo on the left and &ldquosoft&rdquo on the right. Soft lighting includes ambient lighting and shadowing (occlusion) and can be turned on with the command lighting soft or by clicking the Graphics icon .
For setup of the righthand image, see the command file ambient.cxc.
Copyright 2018 Regents of the University of California. All rights reserved.
I would like to thank Christopher A. Hunter, Christopher R. Calladine, Helen M. Berman, Catherine L. Lawson, Zukang Feng, Wilma K. Olson and Harmen J. Bussemaker for their helpful input on the block schematic during its continuous evolution for over two decades. I appreciate Thomas Holder (PyMOL Principal Developer, Schrödinger, Inc.) for writing the DSSR plugin for PyMOL, and for providing insightful comments on the manuscript and the web application interface. I also thank Jessalyn Lu and Yin Yin Lu for proofreading the manuscript, and the user community for feedback.
National Institutes of Health (NIH) [R01GM096889]. Funding for open access charge: National Institutes of Health (NIH) [R01GM096889].
Watch the video: PyMOL Tutorial: Modeling the SARS-CoV-2 RBD Interactions with ACE COVID-19 Coronavirus Proteins (January 2022).