A Bioinformatics How To...
use publicly available free tools to predict protein structure using comparative modeling

For use in the Scripps Graduate Structural Biology Course in Fall 2008

Getting Started

• Start by clicking on Swiss-Model and registering.
• Your password will be emailed to you.
• After you receive your password, login to Swiss-Model
• Download the Swiss-PdbViewer.
• Before, strat your Exercise first click here to review the steps in comparative modeling
• Then read is webpage to get to know number of steps involved in Modeling and Analysis. Now start your Modeling Exercise #1.

Modeling Exercise #1: How accurate is comparative modeling?

In this exercise, we will model the structure of a protein using one of its homologs as a template. Then we will compare the model to the actual structure of the protein.

STEP 1:
The following list of proteins are "easy targets". First try to the modeling for this easy targerts and then try the following seconde list of difficult targets.
Select a target protein from the list below. Then, copy its sequence.



1. The Ligand Binding Domain Of The Vitamin D Nuclear Receptor
2. Shikimate Kinase From Mycobacterium Tuberculosis
3. The Pleckstrin Homology Domain Of Akt1 In Cancer (Akt1-Ph_e17k)
4. Avian Respiratory Complex Ii
5. Bacillus Circulans Strain 251 Cyclodextrin Glycosyltransferase



The following list of proteins are "very difficult targets", that is very difficult to find templates. Their sequence identity ranges from 10-20%.
Select a target protein from the list below and copy its sequence



1. target_protein1.fasta
2. target_protein2.fasta
3. target_protein3.fasta
4. target_protein4.fasta



STEP 2 : Recognition - Selecting a Template

1. Click on the link to the "Template Identification" tool in Swiss-Model.
2. Enter:
   • your email address
   • a title for this part of the project
   • paste your sequence into the large box
   • Choose the PSI-BLAST option, (otherwise use PSI-BLAST (from protein blast, in the "Algorithm" section, select PSI-BLAST))
   • click submit.
3. Once your results are available, examine them. There are several potential template structures for your protein. How should one choose a template?
   
   There are several factors to consider:
   • Which structure has the highest percentage identity with your sequence? Which structure has the lowest percent identity with your sequence?
   • How long is the alignment between your protein and the first potential template shown? How long is the alignment between your protein and the last potential template shown?
   • Which alignment has the highest E-value? Which alignment has the lowest E-value? Which E-value is the best?


So have you selected the best template pdb for your target protein? What is it?

Your Protein Sequence	Your Template PDB
target_Protein1	???
target_Protein2	???
target_Protein3	???
target_Protein4	???




STEP 3 : Alignment

• Examine the alignment of your protein with the template you have chosen.
• We could either use this alignment for the modeling or we could refine the alignment by using different parameters to construct the alignment.
• If you would like an extra challenge follow the steps below to refine the alignment.
• If not, write down the pdb code for your template (the pdb code is 5 characters long, it starts with a number which is followed by three letter or numbers, the fifth character is usually a letter, for example 1abcA) and begin step #4.

Challenge

• To recognize the template, we used the default settings in "Template Identification" tool, which used the blosum62 substitution matrix.
• To align closely related seqeunces, a blosum matrix with a high number should be used.
• Conversely, when aligning distantly related sequences, a blosum matrix with a low number should be used.
• The "Template Identification" tool will let you choose to use blosum80, blosum62, or blosum45.
• What is the percentage of sequence identity between your target (your sequence) and your template? Which blosum matrix should you use?

There is one or more alignment parameter to consider.

• When two proteins are aligned, there are many possible ways to align the protein.
• The best alignment has the highest score, addtioally we should consider the following:
• More number of aligned residues that are identical or similar between your query sequence and template will increase the score.
• Gaps in the alignment decrease the score. Each gap decreases the score with a penalty for opening the gap (this is called "G" in the "Template Identification" tool) and a per-residue cost for extending the gap after the first residue (this is called "E" in the "Template Identification" tool).
• A high "G" tends to decrease the number of gaps in the alignment. A high "E" tends to decrease the length of the gaps.
• So, choose the appropriate blosum matrix and gap penalties, and realign your target and template using the "Template Idenitifcation" tool.
• Copy the alignment of your target and your template. Also, write down the pdb code for your template (the pdb code is 5 characters long, it starts with a number which is followed by three letter or numbers, the fifth character is usually a letter, for example 1abcA).




STEP 4 : Modeling

• In the "Alignment Formater" below, enter the name of your target, the name of your template, and your alignment. Then click on "SEND".
  
  

  Alignment Formater

  
  

  Name of Target: 
  
  Name of Template: 
  
  Alignment: 
  
  
   
  

• Copy the output from the "Alignment Formater".
• Click on the link to SwissModel Alignment Model your protein. Enter your email address and a title for this part of the project. Paste the output from the "Alignment Formater" into the large box. Then click "submit alignment."
• The next webpage from Swiss Model will ask you to "Please select sequences from your alignment." Next to target sequence, use the pull-down menu to select the name you gave your target sequence. Next to template sequence, use the pull-down menu to select the name you gave to the template sequence. Also, please make sure that the pdb code is correct. Then click "submit alignment".
• Once you have your results, click on, "download model: as pdb" to save your model as a pdb file. Make sure that you know the name and location of your pdb file as we will use it later. Keep the Swiss model window with your results open as we will use this window in the next step.



STEP 5 : Model Analysis

Using Swiss-Pdb Viewer for model visualization and analysis

• >>To View your Model in Swiss-Pdb Viewer:

  Now, that you have a model, how can we tell if it is any good? Let's look at your model using some tools in Swiss-Pdb Viewer.
  
  
  1. Click on the click from Swiss Model to "display model: as DeepView project".
  2. This may open a window which asks if you would like to open the file or save it to disk. If it does, open the file with spdbv.exe.
  3. If a window called "About Swiss-PdbViewer..." opens, close it.
  4. Now, you should have two windows open: one with a toolbar and several pull-down menus and another that is a protein structure viewer.

• >>Moving Your Model for Easier Viewing:

  Now that you have opened your model, we'll learn how to move it around the screen. There are four buttons in the toolbar that are important for viewing your model.
  
  
  5. Click the button on the far left, which looks like a window with a structure in the middle and three arrows pointing to it. You should notice that the structure is now centered and zoomed in to a size appropriate to the size of your structure-viewing window. This button is the center and zoom button.
  6. Now, try clicking on the second button from the left in the toolbar, which looks like a hand. Go to the structure-viewing window and drag the structure around the screen. This button is the translate button.
  7. Click on the third button from the left in the toolbar, which looks like two boxes: one solid box and one with dashed lines. Go to the structure-viewing windows and drag the mouse around the screen. This is the zoom button.
  8. Click on the fourth button from the left in the toolbar, which looks like a circular arrow. Go to the structure-viewing window and drag the mouse around the screen. This is the rotate button.

• >>Viewing Only Your Model:

  Now that you know how to move the model around the screen, let's start the analysis!
  
  
  9. In the window with the toolbar and pull-down menus, click on the "Window" menu and select "Layers Infos". This will open a new window called "Layers Infos". In Swiss-Pdb Viewer, each model or structure is displayed in it own layer. In the "Layers Infos" window, there should be information about two layers: one called "TARGET" and one named for the template structure you used to build your model.
  10. Go back to the window with the toolbar and pull-down menus. Click on the "Color" menu and select "by Layer". Now, your template is blue. The wireframe display of your model is yellow and a ribbon diagram of your model is green and perhaps some other colors as well.
  11. Go to the "Layers Infos" window. Next to the name of each layer, there is a column labeled "vis". Click on the "v" next to your template in this "vis" column. Your blue template should disappear.
  12. In the "Layers Infos" window, click on "TARGET". Now the word "TARGET" should be red.
  13. In the window with the toolbar and pull-down menus, click on the "Windows" menu and select the "control panel". A new wondow should open.
  14. In the left-most column of this new window, all of the residues in your model are listed. To the right of this column is another column that is labeled "show". Press "shift" while clicking on ones of the "v"'s in this second column from the left. Now only the ribbon diagram of your model should remain in the structure-viewing window.

• >>Geometry-based Tools to Gauge Model Quality:

  Now we'll use geometry-based tools to gauge the quality of your model.
  
  
  15. In the window with the toolbar and pull-down menus, click on the "Select" menu and select "aa Making Clashes". Press "enter". This will display amino acids that are in clashes with other amino acids. If no amino acids are displayed, there are no amino acids in clashes with other amino acids in your model.
      
      How many amino acids are clashing with others? None, some, or many?
      
      
  16. Go back to the window with the toolbar and pull-down menus. Click on the "Select" menu and select "aa Making Clashes with Backbone". Press "enter". This will display amino acids that are in clashes with the backbone of your model. If no amino acids are displayed, there are no amino acids in clashes with the backbone in your model.
      
      How many amino acids are clashing with the backbone? None, some, or many?
      
      
  17. Go back to the window with the toolbar and pull-down menus. Click on the "Select" menu and select "Sidechains lacking Proper H-bonds". Press "enter". This will display sidechainsthat lack proper hydrogen bonding.
      
      How many sidechains lack proper hydrogen bonding? None, some, or many?

• >>Empirical energy-based Tools to Gauge Model Quality:

  We can also use tools based on empirical energy to gauge the model quality.
  
  
  18. In the window with the toolbar and pull-down menus, click on the "Window" menu and select "Alignment". A new window will open that shows the sequence alignment of your target protein and your template.
  19. In this window, there is a row that is labeled, "TARGET". Click on the word, "Target", so that it is red.
  20. To the left of this label is a red question mark. In between the label and the question mark, there is a white triangle that points towards the "TARGET" label. Click on this triangle. This increases the height of the "Alignment" window. Now above the alignment, there is a line and a curve displayed.
  21. In the window with the toolbar and the pull-down menus, click on the "SwissModel" menu and select "Auto Color by Threading Energy".
  22. In the "Alignment" window, click on the word, "smooth". A new window will open.
  23. In this window, set the "nb of aa to average" to 1.000 and click "ok".
  24. In the "Alignment" window, click on the "E=" text, which is above "smooth". This will recalculate the threading energy for your model. The line, in the "Alignment" window, shows where the threading energy is zero. The curve shows the threading energy per residue (actually the average energy of each residue plus the one residue towards the N-terminus and one residue towards the C-terminus) of your model. A good model should have an average threading energy of 0. A long stretch of residues with threading energy higher than zero and colored red or orange may indicate that this part of the model is not accurate.

• >>Other Considerations:

  Another consideration is the function (activity, process) of your protein.
  • Does the template have the same or a similar function?
  • If so, how good is the alignment of the active site or residues involved in this function?
  • Are these residues conserved?

STEP 6 : Advanced Model Analysis

• Comparing the Model to the Actual Structure of the Protein

  The proteins we are using in this exercise actually have known structures. So we can compare your model to the actual structure.
  • To compare your model to the actual structure of your protein:
  • Then, click on the link to FATCAT.
  • Enter your email address.
  • Upload the 1st structure, that is the pdb file containing your model and name this protein, "modelProteinName".
  
  
  

  Comparing the "easy targerts" Models to the Actual Structure of the Protein, firts download the actual pdb structres from the PDB Database

  Your Protein Model	PDB code	Chain
  The Ligand Binding Domain Of The Vitamin D Nuclear Receptor	2hcd	A
  Shikimate Kinase From Mycobacterium Tuberculosis	2dft	C
  The Pleckstrin Homology Domain Of Akt1 In Cancer (Akt1-Ph_e17k)	1p6s	A
  Avian Respiratory Complex Ii	1nek	B
  Bacillus Circulans Strain 251 Cyclodextrin Glycosyltransferase	1cyg	Since this protein only has one chain, leave this field blank in FATCAT.

  
  

  Comparing the "difficult targerts" Models to the Actual Structure of the Protein

  Your Protein Model	Download Actual Structure files	Chain
  model_protein1	GS13115C_protein1.pdb	A
  model_protein2	MG14668A_protein2.pdb	A
  model_protein3	MG14474A_protein3.pdb	A
  model_protein4	MG14461A_protein4.pdb	A

  
  

• Enter these for the 2nd structure and name this 2nd structure:

  • First download the the actual structures in the column "Download PDB coordinate files" corresponding to your model protein from the above table.
  • Upload the pdb coordinates of the the actual structure corresponding to your model protein, and name this protein, "ActualStructureName".
  • Click on "Send Request".
  • While you wait for your results, check that your computer has either Chime or Jmol installed.
  • If your computer does not have either installed, click on either Chime or Jmol to download one of these programs.

• Once you get your results, how similar is your model to the actual structure of your protein?

  • What is the RMSD?
  • Examine the superimposition of your model and the actual structure.
  • Which parts of your model look the most similar to the actual structure?
  • Which parts of your model are clearly different than the actual structure?

• Discuss your results with your classmates.

  • How similar were their targets and template in terms of percent sequence identity?
  • What is the RMSD between their model and the actual structure?
  • Which parts of their model was the most and least accurate?

In this exercise, you will model the structure of a protein and gauge the quality of the model. This protein does NOT have a solved structure.

1. Either select a target protein from the list below and copy its sequence or use the sequence of your favorite protein.
1. 3BP2_MOUSE SH3 domain-binding protein 2 (3BP-2)
2. CHICK Xenobiotic receptor

2. Recognition

   Using the "Template Identification" tool in Swiss-Model and what you learned about template recognition in the last exercise, select a template for your protein.

3. Alignment

   Examine your alignment. If you think it is good, copy it. If not, use the suggestions in the challenge portion of the "Alignment" section in the last modeling exercise to refine the alignment of your protein to its template.

4. Modeling

   Use the Alignment Mode in Swiss-Model to model your target protein.

5. Model Analysis

   Use the instructions from the last exercise to:
   • View your model
   • Use geometry-based tools to gauge the quality of your model
   • Use empirical energy-based tools to gauge the quality of your model
   • Use functional considerations to gauge the quality of your model

Programs and Webservers Used in These Exercises

• Swiss-Model
• Swiss Pdb Viewer (also called DeepView)
• Chime
• Jmol
• FATCAT

You can also try advanced modeling tools available from The Joint Center for Molecular Modeling (JCMM) - Server Links