Possible Molecular Mechanics Projects
Pick any one of the following projects (feel free to do more, but I only want one report). Some of these have been done, some not.
- The purpose here is to develop your skill in using the programs, and to practice defining a problem in such a way that computation can provide an answer.
- You can define the problem pretty much to suit yourself. See what you can learn.
- Write a one-page report providing both your answer(s) and the computational methods you employed and why you selected them.
- Feel free to consult the literature to see what others have to say, but cite any literature you use.
- If you would like to develop a small project related to your research, consult with me before beginning.
- The PCBs are a family of chlorinated biphenyls that are claimed to have all sorts of evil properties, none of which have been proven for humans. Nonetheless, their use is banned. Of particular interest is 2,3,4,3',4'-pentachlorobiphenyl, which is referred to by biologists as a "coplanar biphenyl", and argued as a consequence of its coplanarity to have toxicity comparable to dioxins. Is it coplanar? If not, what would be the energetic cost of making it coplanar? What happens to the coplanarity if you remove some of the chlorines?
- Many years ago, the great Roger Adams demonstrated the chirality of o, o'- substituted biphenyls, such as the PCBs above, and those shown below (X = Br or I).
HERE is a page discussing the chirality of these species.
Subsequently, in the first known example of a molecular mechanics calculation, the equally great Frank Westheimer, working with physical models, a protractor, and a ruler, estimated the barrier to racemization of these compounds as 17.3 and 21.0 kcal/mol, respectively. He then made a suggestion about which way the rings rotate during racemization. How good were his numbers, and which way do the rings rotate?
- The structure below has been suggested for a homogeneous, enantioselective, hydrogenation catalyst.
If the catalyst is enantioselective, it must itself be chiral. Is it?
- One can imagine that cyclic ketones could exist in two families of conformations: one with the carbonyl "outside" the ring, and one with it "inside". These structures are shown below for cyclodecanone. Clearly, small rings would not permit the "inside" conformation, whereas larger rings might. At what ring size does the "inside" conformation become possible? Can you find a ring in which "inside" is preferred? This might be a bridged structure, or simply a large ring. Explore.
- As shown below, when 2-methylcyclohexanone reacts with pyrrolidine,
the enamine is preferentially formed with the double bond toward the unsubstituted carbon. We know that normally, the more substituted double bond is the more stable, but it has been suggested that conjugation of the nitrogen unshared pair with the double bond causes the pyrrolidine ring to be coplanar with the double bond. This in turn would lead to a serious steric interaction if the double bond were on the methyl-substituted side. Is this so? Does the outcome depend upon ring size? If so, is there a switchover point?
- Copper (II) complexes of N-methylamino acids have the general structure shown below for N-methylalanine.
Are the ligands around copper placed square planar or tetrahedral? With two stereogenic centers, this kind of complex can exist in diastereomeric forms. What is the energy difference between them (this is important because such complexes are sometimes used to resolve racemic amino acids).
- Obtain the crystal structure of crambin from the Protein Data Bank (1crn). Minimize it. Compare your minimized structure to the crystal structure. Determine the rms difference. How does the secondary structure change upon minimization?
- Sorry, folks! this one won't work. PCModel doesn't write .pdb files that can be read properly at the PDBSum website. If you've been working on this one, just make pictures of the two best structures and hand those in. Tuftsin is the tetrapeptide Thr-Lys-Pro-Arg. It is a signaling peptide, involved in our defense mechanisms against bacteria. Use PCModel to search the conformation space of the backbone angles, and minimize the two lowest energy structures you find. How do your structures compare with those found by O'Connor, et al, J. Med. Chem., 1992, 35, 2870? Save your structures in .pdb format, and upload one of them to the PDBSum website for a full analysis. Use the "Generate" button on the left-hand vertical menu for uploading.
When you receive your email with the URL for your results, click on the Ramachandran diagram (headed Procheck) and then click on Generate Full Procheck Analysis. From the page that pops up save the Main Ramachandran plot as a .pdf and the Distorted Geometry .pdf, and submit them with your report.
- Cyclodextrins are structures formed by cyclizing relatively short chains of glucose units in an a 1,4-linkage. Alpha-cyclodextrin is six units, b- is seven units, and so on. They are created by the action of various bacteria on starch.
Recently, these molecules have attracted attention as drug delivery agents because of their ability to bind small molecules inside the ring, protecting them from hydrolysis in the stomach. This project involves figuring out how large the molecular cavity is.
Download the structure 1vfm from the Protein DataBank (www.rcsb.org). It is the structure of a protein complexed with two molecules of a-cyclodextrin. Cut out one of the cyclodextrins from the file, retaining the PDB format, and read it into PCModel. Optimize the structure. Is the optimized conformation any different from the one bound to the protein? What is the approximate diameter of the "hole" in the center?