Ideas for Modeling Projects

Following are a few ideas for modeling projects. You could either do one of these or come up with an idea of your own.

Remember:

1. Homology Modeling of Chemokine Structure
Model the structure of the chemokine eotaxin based on sequence alignment with the CC-chemokine MCP-1 and the average minimized NMR structure (monomer only) of MCP-1 (PDB file: 1dom). Compare the structure of your model, the NMR structure, and the x-ray structure to locate different structural differences. Identify whether the differences are due to energetic differences between the structures. To investigate these differences, use Discover/Constraint/template_force to force one structure to conform to another structure.
(This project is of interest to the Stone lab)

2. Prediction of Cross-Linking Chemistry
One common method to probe dimer formation under various solution conditions is to add a chemical cross-linking reagent. Cross-linking reagents with a wide variety of chemical reactivities are available (see Pierce catalog). Using the known 3D solution structures of the MCP-1, MIP-1b, and RANTES dimers, predict the products of reaction with a variety of cross-linking reagents (e.g. no reaction, specific reactivity, random reactivity, etc.).
(This project is of interest to the Stone lab)

3. Alignment of a Protein Family: Prediction of Important Residues
Do a database search for members of the chemokine family (probably around 30 members now) and align these sequences. Use the known structures of CC and CXC chemokines to display in 3D the residues that are most highly conserved and predict the reasons for the conservation of these residues.
(This project is of interest to the Stone lab)

4. Interactions of Vancomycin with its Target
Vancomycin is an antibiotic that binds specifically to the D-Ala-D-Ala C-terminal dipeptide of nascent bacterial cell walls. Bacteria that are resistant to vancomycin produce the alternative D-Ala-D-lactate in their cell walls. Build vancomycin and do MD to find the most stable structures. Model the binding of N-acetyl-D-Ala-D-Ala and N-acetyl-D-Ala-D-lactate to vancomycin and suggest why D-Ala-D-lactate overcomes vancomycin inhibition of cell wall growth. Try changing the structure of vancomycin to increase its affinity for D-Ala-D-lactate.

5. Electrostatic Mapping of the Thioredoxin Active Site
Thioredoxin is a protein that catalyzes disulfide exchange reactions. It contains two Cys residues at its active site, one of which has a pKa several pH units lower than normal (resulting in some special chemical properties). Use the known structure of E. coli thioredoxin to predict the reason(s) for this unusually low pKa. Then test your hypotheses by modeling some thioredoxin mutants and predicting the pKa values of the Cys residues.

6. Protein Structure Calculation using Distance Restraints
For a small protein of known structure, make a list of proton-proton distances that might be detected by NMR NOESY experiments. Use this list (or a subset of this list) along with the primary sequence of the protein to do a distance geometry calculation then a molecular dynamics run to calculate the structure of the protein. Hopefully you should get back the structure that you started with.

7. Electrostatic Mapping of DNA-Binding Proteins
Most DNA binding proteins are positively charged in the region that interacts with the phosphodiester backbone. Using the known structures of several homologous DNA-binding proteins (e.g. helix-turn-helix domains), calculate and compare electrostatic potential maps of the proteins to predict the regions important for DNA binding. Also, map the electrostatic potential surface of a DNA double helix and look for complementarity between the two. Extensions could involve "manually" docking one of the proteins with its DNA recognition sequence and predicting the residues important for base-specific recognition.

8. De novo Modeling of Aptamer Structures
Based on the sequence and predicted secondary structure of an RNA aptamer, use molecular dynamics (with hydrogen bonding constraints) to predict the tertiary structure of the aptamer.
(This project is of interest to the Ellington lab- See Andy to discuss details).

9. Docking Aptamer to Protein
Dock the known structure of an RNA aptamer to the known structure of its target protein.
(This project is of interest to the Ellington lab- See Andy to discuss details).

10. Mutational Analysis of Protein-DNA Interactions
Based on the known structure of a protein-DNA complex (e.g. zif268-DNA) predict changes in the protein sequence that would allow specific recognition of different DNA sequences. See if you can come up with a protein-DNA recognition code. The zinc finger family would be a good example.

Delete parts or all of selected side chains of the protein. Design natural or unnatural side chains that would bind DNA more tightly. Use LUDI (ligand design) to have InsightII automatically design new side chains for you. Modify the DNA sequence and and evaluate of your new side chains can bind to the new DNA sequence. Can you develop a modified side chain or several modified side chains that SPECIFICALLY binds to ONE DNA sequence???

11. Phosphoribosyl Transferases - Homology Modeling and Substrate Binding
The Taylor lab is working with the enzyme APRT and have cloned two new genes in this family. One idea is to build a homology based model of one of the new proteins. A second idea is to dock the substrate to the enzyme. They have some mutagenesis data which would help in identification of the active site.
(This project is of interest to the Taylor lab- See Milton to discuss details).

12. Structure Modeling of Interferons
Amgen produces a form of interferon with a "consensus" sequence and with higher specific activity than the natural interferon. Whether there is a significant difference between these proteins is currently the subject of a legal battle. The Taylor lab is interested in modeling the structures of these proteins and deciding whether they differ significantly. Modeling would presumably be based on homology to some known structure(s).
(This project is of interest to the Taylor lab- See Milton to discuss details).

13. Modeling an Active IL-5 monomer
Interleukin-5 (IL-5) is a dimer with a known structure in which the two monomers are somewhat intertwined, whereas other closely related cytokines are monomers (four helix bundle structures). Based on the known structures and on sequence alignments, design a variant of IL-5 which is likely to be both monomeric and active and a variant of one of the other cytokines that is likely to be both dimeric and active.

14. Molecular Dynamics of an a-Helix
Build an a-helix and do an extended molecular dynamics run then analyze the motions in the helix. Repeat using different temperature and look for variations in observed motions.

15. Docking your Favorite Small Ligand and Big Protein
Examples that come to mind are: proteases and their inhibitors (Roush lab); phosphotyrosine phosphatases to their ligands (Widlanski lab); odorant binding proteins to their ligands (Novotny lab); triosephosphate isomerase to inhibitors (Stone lab). Start with a known protein structure. Try a few ligands and make some predictions about specificity.

Back to  |  C687 Summer 1997  |  Courses & Instruction  |  MolViz Home  |
Send comments to chemvis@indiana.edu
Last updated: 01/23/2001