##################################### Bio 5476 Protocol for Lab Exercise #1 ##################################### (1) Login to your account using the login name and password distributed in class. This login/password will work on any machine in the CCB teaching lab. Your accounts are set to use the "bash" shell. You may wish to drag some heavily used programs to your Dock. For example, the Unix terminal (ie, /Applications/Utilities/Terminal.app) and a basic editor (ie, /Applications/TextEdit.app). Also, you should create a .bashrc file in your Home directory and have it contain the command "set nohup". This command should allow submitted jobs to continue to run after you log out. (2) While in your Home directory, copy the TINKER programs, sample molecules and force field parameter files via the command "cp -r /Applications/MolecularTools/tinker ." issued in a Terminal window. Alternatively, just drag the "tinker" folder under /Applications/MolecularTools to your home directory. (3) After checking to make sure FFE is available, try opening some of the sample molecules in the ~/tinker/test directory. Practice rotating, coloring, resizing, changing the representation, etc. Also try launching some of the TINKER programs such as ANALYZE or SPACEFILL on a sample molecule. You can drag the "Force Field Explorer.app" icon to your Dock to make it easier to launch the GUI in the future. (4) Exit FFE. Then make a /work directory in your home area, change to this new /work directory, and relaunch FFE from the /work area. (5) Now run the PROTEIN program from the Modeling Commands panel of FFE and build the sequence ASP-TYR-NME. This is very similar to the structure of the sweetener aspartame which is ASP-TYR-MeEster. Use the OPLS-AA force field parameters in ~/tinker/params/oplsaa.prm to assign the atom types. (6) Once the peptide is generated, try Cartesian minimization using each of the three minimization algorithms build into TINKER, ie, the programs MINIMIZE, OPTIMIZE and NEWTON. These programs implement algorithms of the limited-memory BFGS (conjugate gradient-like), variable metric and truncated Newton classes, respectively. See how many iterations are required by each method to reduce the RMS gradient to the default value of 0.01 kcal/mol/Ang per atom. After each minimization, you will need to remove the minimized structure to enable the next minimization to start from the original "protein"-built structure. Finally, repeat the three minimizations using a convergence criterion of 0.0001 kcal/mol/Ang. (7) Repeat the above using minimization in torsional space. The TINKER programs to try are MINIROT and OPTIROT. Again, how many iterations are required. Which seems faster in terms of CPU time, Cartesian or torsional minimization? (8) Next create another peptide ACE-VAL-VAL-NME via the PROTEIN program. Use the TINKER SCAN program to find all the local minima for this peptide. This calculation may take some time, so you should submit it in the background. Since this SCAN is being done in the "gas phase", what feature would you expect to find in all of the lowest energy conformations of the peptide? Use the ARCHIVE program to copy all the minima found by SCAN into a single .arc file. Play this back as a movie using FFE. Examine some of the lowest and highest energy structures, to see if they have feature in common. (9) Delete all open structures in FFE. Now go to the Protein Data Bank web site (www.rcsb.org) and get the PDB file containing the structure of the small plant-seed protein crambin. One possible PDB code for crambin is "1CRN", but more recent structures are also available. (10) Use the TINKER PDBXYZ program to convert the PDB file format to the Cartesian .xyz format using internally by TINKER. Choose the Amber ff99 force field parameters. (11) Turn on the GB/SA continuum solvation model using the Keyword Editor, and then run a short minimization of crambin using the MINIMIZE program with a target RMS gradient of 1.0 kcal/mol/Ang. (12) Next start a dynamics trajectory for the GB/SA solvated crambin using the TINKER DYNAMIC program. Use a time step of 2.0 fs, and set it to run a total of 500000 steps. This will correspond to a total run of 2 x 500000 = 1000000 fs = 1.0 ns of simulation trajectory. Choose to target the system thermostat to 298K, which is about 25 C. Save (or "dump") trajectory coordinates every 1 ps, which would be every 500 time steps. Also set the trajectory to be saved directly to an archive file in the Keyword Editor. Before starting the run, manually set the random number seed to some random large number. (13) Once the MD is done (which may be well after lab is finished....), run the original PDB file and the MD archive file through the SUPERPOSE program and generate a plot of the alpha carbon RMSD of the PDB structure vs. the MD structures as a function of the time along the MD trajectory. (14) Turn in a brief writeup of the lab, with the information you collected comparing minimization methods, analyzing the low energy conformations of the ACE-VAL-VAL-NME peptide, and performing the crambin MD run.