FDiscover

4       Tutorial

This section describes several lessons on using FDiscover commands in the Insight environment and in standalone mode.

Pilot online tutorials

Then, from the Open Tutorial window, select Discover tutorials (in release 97.0) or Molecular Mechanics Tutorials and then Discover Tutorials (in release 4.0.0) and choose from the list of available lessons:

Lesson 1 Minimizing and performing molecular dynamics
Lesson 3 Using pseudoatoms
Lesson 4 Consensus dynamics/minimization

(Lessons 2, 5, and 6 are included in this section.) You can access the Open Tutorial window at any time by clicking the Open File button in the lower left corner of the Pilot window.

For a more complete description of Pilot and its use, click the on-screen help button in the Pilot interface or refer to the Insight II User Guide.

Overview of tutorial lessons

• Retrieving a model from a folder.

• Optimizing the conformation of a model with FDiscover.

• Performing dynamics on a model with FDiscover.

• Saving the model in a folder.

Note: This lesson consumes large amounts of CPU time and disk space. It may take days to complete on a Silicon Graphics Personal IRIS. You must have at least 12 MB of free disk space.

The calculation of the solvation free-energy difference entails using a thermodynamics cycle and calculating relative free energies. FDiscover uses the finite difference thermodynamic integration (FDTI) method to calculate the free energy difference between two chemically distinct models (in this case, methanol and ethane). FDTI is a thermodynamic integration scheme that uses the perturbation method to evaluate the integrand.

In Lesson 2, you create methanol and ethane models, fix the potentials and charges for your models, set up the Discover job for the gas phase, build the hydrated system and set up the Discover job for the aqueous phase, run the jobs, and interpret the results. The topics covered in this lesson are:

• Defining the relationship between initial and final states of a model.

• Defining parameters for a relative free energy job.

• Writing out Discover input files.

• Solvation of a model with periodic boundary conditions.

• Submitting a Discover job and interpreting the results.

• Building a model with the Builder module.

• Defining pseudoatoms with the Pseudo_Atom/Define command.

• Defining distance restraints with the Constraint/GenericDis command.

• Setting up an FDiscover simulation.

• Monitoring the restrained distances during the Discover simulation.

Consensus restraints have several uses. One example is the determination of structural similarities among a set of homologous models--perhaps of several similar compounds that bind to a particular receptor. If you hypothesize that an apparently homologous region of the compounds is responsible for the binding, then you would want to find a configuration for this region that is compatible with relatively low-energy conformations for all the models. By applying consensus restraints to the homologous region of each model, you can use dynamics followed by minimization to find a likely binding configuration.

In this tutorial, you will consider four different inhibitors of angiotensin-converting enzyme (ACE): captopril, SA446, oxocaptopril, and butcaptopril. The object of this lesson is to use consensus restraints to find a set of structures for the inhibitors in which the proposed pharmacophoric region has the same conformation.

The topics covered in this lesson are:

• Generating and editing input files for the FDiscover program

• Reading in models from a preexisting .psv file.

• How Discover can be used to examine the structure of a peptide without using Insight.

• Given two of the three required input files (.car and .mdf), how to instruct Discover what to do by specifying commands in the .inp (command input) file.

• Performing dynamics and minimization in the standalone mode.

Lesson 2: Relative free energy

To begin this lesson, issue the command insightII at the system prompt, from a directory in which you have write permission.

It takes a few moments for the Insight program to load.

2. Creating a methane and an ethane

Go to the Module pulldown (click the MSI logo) and select Builder.

There are several ways to construct the desired molecules. You may want to read ahead and then use, say, the Sketch pulldown instead to perform Steps 2 and 3.

Also note that the Builder interface is somewhat different in Insight 96.0 and 4.0.0.

Select the Fragment/Get command. After the Get Fragment parameter block appears, toggle To_Modify_Bond to off (it changes from yellow to the background color), so that you will remain in this parameter block after executing the command. Then go to the Fragment Window and pick any atom in the methyl fragment. Change the name of the model by entering MEOH in the Get Molecule parameter box, then select Execute (in the Get Fragment parameter block). After the methane model appears in the center of the screen, move it upwards a little in the display area. Now pick the ethyl fragment from the Fragment Window, enter ETH in the Get Molecule parameter box, and select Execute again.

3. Changing the methane to methanol

Select the Modify/Bond command. After the parameter block appears, toggle the Fragment_Window parameter on. Go to the Fragment Window and pick a hydrogen of the hydroxy fragment, followed by a hydrogen of the methane model.

An OH group is attached to the methane model, to form methanol.

4. Correcting the potentials and charges

Select the Forcefield/Potentials command. When the parameter block appears, choose Fix as the options for the Potential Action and Partial Chg Action parameters. Then pick any atom in the methanol model. The command executes automatically. Select the Forcefield/Potentials command again, set the same action parameters to Fix, and pick any atom in the ethane model.

The value of the Formal Chg Action parameter is important in Apex-3D calculations and has no effect in the Discover program.

5. Defining the warped atoms

The calculation of free energy involves the warping of atoms from one chemical state to another. The starting state is referred to as the A state, and the final state is referred to as the B state. Warping is done by giving an atom some of its chemical characteristics from the A state and some from the B state.

In the beginning of the simulation, a warped atom exhibits almost all A state characteristics, while in the end it exhibits almost all B state characteristics. The characteristics being warped include all the potential function terms (bond, nonbond, charge, etc.). In certain simulations, such as this one, the number of atoms in state A and state B is different. An atom that is not present in state A, but is present in state B, is warped from null to a real atom. A null atom is merely a placeholder in the molecule's topology. It does not contribute any energy to the model. An atom that exists in the A state, but not in the B state, is warped to null.

To aid in the definition of the warped atoms, select the Molecule/Label command from the upper menu bar. When the parameter block appears, type an * in the Molecule Spec parameter box, then press <Enter> or select Execute.

This causes both the ethane and methanol models to be labeled with their atom names. (The * is a wildcard and means "all objects" when used in this manner.)

From the Module pulldown (the MSI logo), select Discover. When its pulldowns appear on the lower menu bar, select the Constraint/Warp command.

In this example, you are warping methanol to ethane. This involves warping an oxygen to a carbon, and warping from null to two additional hydrogens, since ethane has two more hydrogens than methanol.

It is important to note that the partial charges on an atom are determined by what that atom is bonded to. The atom pairs that are changing only in charge need to be warped in addition to those pairs that are changing both in potential atom type and charge. For this example, this involves warping the carbon of the unchanging methyl group of both models. This also involves warping the hydrogen bonded to oxygen in methanol to a hydrogen bonded to carbon in ethane. The State A and State B molecules are shown in Figure 1.

To establish the pairs of atoms being warped, make sure that the Activation parameter is set to Add and Warp_Type to Atom_Pair. Connect to each model in turn (with the Connect Object button) and move them so that they are oriented approximately as shown in Figure 1. To perform the warping of carbon C1 in State A to State B, pick C1 in methanol as the A_Atom, followed by C1 in ethane as the B_Atom.

Figure 1 . State A model (methanol) and State B model (ethane) for tutorial on relative free energy

Picking the second atom triggers the command, so you do not need to select Execute. As a check, the A_Atom parameter box should now be filled with MEOH:1:C1, and the B_Atom parameter box should now be filled with ETH:1:C1.

To perform the warping of oxygen OH in State A to carbon C2 in State B, pick OH in the methanol model followed by C2 in the ethane model.

Picking the second atom triggers the command; you do not need to select Execute.

To perform the warping of the hydrogen HH in State A to hydrogen H23 in State B, pick HH in the methanol model, then H23 in the ethane model.

All that remains to be warped are the two additional hydrogens bonded to the C2.

Set the Warp_Type parameter to From_Null.

The B_Atom parameter is automatically highlighted (depending on the settings in the Session/Cmd_Display command, the A_Atom parameter box may disappear).

Pick H21 in ethane, to warp the hydrogen H21 from null to a hydrogen.

Picking the atom automatically triggers the command.

Pick the H22 in ethane, to warp the hydrogen H22 from null to a hydrogen.

6. Setting up the free energy simulation

Now, select the Parameters/Relative command. When the parameter block appears, check to make sure the following parameters are set: Lower_lambda: 0.00 (the lower limit of the free energy integration) Upper_lambda: 1.00 (the upper limit of the free energy integration) Dlambda: 0.005 (the step size of lambda) Quadrature Points: 6 (the number of Gauss-Legendre quadrature points) Sampling: 10 (the frequency at which exp (-H/kBT) is sampled).

If these values are set, you do not have to select Execute, since they are the defaults.

The next step is to set the dynamics conditions.

Select the Parameters/Dynamics command. When the parameter block appears, change the number of equilibration steps at each lambda point to 10,000 and the number of dynamics steps for data collection at each lambda point to 10,000, by entering the following values Equilibration = 10000 Steps = 10000 History = 500 Charges = on (yellow) Cross = off (background color) Morse = off Temperature = 300.00 Time Step = 1.00 Constant_Pressure = off Select Execute.

Note that the number of steps needed for equilibration and each quadrature point varies from system to system. Here, 10,000 is an approximate number, but you should always plot the batch average of exp (-H(+) / kBT) versus time (Column 5 versus Column 2 of your .tot file) for each lambda to assure convergence. This number should fluctuate about a constant over time.

Select the Parameters/Variables command. When the parameter block appears, toggle the Cutoff parameter to on and enter 9.8 as the Cut_off value. Select Execute.

Note that the cutoff parameters are not necessary for the gas phase, but are specified now for the aqueous phase.

To write out the files needed for starting this job, select the Run/Run command. Make sure that the Object Name is ETH (type it in or select ETH from the list of Assem/Mol Names in the value-aid, if needed).

Even though the beginning model is methanol (State A, where = 0), you select ethane to write out the .car and .mdf files, because it has the greater number of atoms. The coordinate and molecular data files written out will be a combination of both models. The atom and residue names and coordinates are taken from the model with the greater number of atoms (ethane, in this case). This is done to guarantee that every atom has a real name and coordinate, even though some may start out invisible (null atoms). The potential atom types are always taken from the beginning model (methanol, in this case).

Now set the Computation Mode to Command File. Make sure that Auto Option is set to Add_Auto, that Run_Minimization is off, and that Run_Dynamics is on. Select Execute to write out the .inp, .car, and .mdf files for Discover.

However, you need to modify the input files before starting the job. This will be done in Step 10.

7. Clearing all existing constraints

You must now clear the existing warped atom definitions.

Select the Constraint/Warp command. When the parameter block appears, set Activation to Clear and select Execute.

8. Building the hydrated free energy system

Delete all objects. Do this with the Object/Delete command. Enter an * for the Object Name and select Execute. Read the newly created .car file back in with the File/Import (or Molecule/Get) command: Select the File_Name parameter if it is not already active, then choose eth0.car from the Files value-aid. You can scroll the list in the value-aid, if necessary, by using the scroller on its right side. Select Execute.

This redisplays the ethane model. (It may be recolored by entering colat on the command line near the bottom of the window.) The name of the model is now ETH0, not ETH.

The next step in the free energy evaluation is the creation of the solvated model, which will be compared with the free energy calculated in the gas phase.

You need to create a water box with a volume of 15.6 ų, with the solute at the center. To do this you need to move the center of your solute (ethane model) to the center of the box.=

Select the Transform/Move command. The Object Name parameter should already be filled in with ETH0. Under the Move Values parameter, activate the parameter box for X by clicking in it. Enter 7.8 and press <Enter>. Also enter 7.8 for both Y and Z. Press <Enter> or select Execute. To write out the structure at these coordinates, select the File/Export (or Molecule/Put) command. When the parameter block appears, make sure the Put_File_Type is Biosym, select ETH0 from the Assembly/Molecules value-aid or by clicking the model, type eth1 in the File Name box, and toggle Transformed to on. Select Execute. Delete the current model by using the Object/Delete command. Then read in the model with the transformed coordinates with the File/Import (or Molecule/Get) command. Fill in the File_Name parameter by selecting eth1.car from the value-aid. Select Execute.

Now you need to set up the periodic boundary conditions for this box.

Select the Assembly/Cell command. Set the Object Name to ETH1. Enter the following parameters in the Lengths parameter boxes: a: 15.6, b: 15.6, c: 15.6. In the Angles parameter box make sure the parameters have the following values: a: 90.00, b: 90.00, c: 90.00. Leave the Table Number set to 1 and select Execute. Solvation is accomplished by selecting the Assembly/Soak command. Change the Method to PBC and select Execute.

9. Writing the Discover files

To write out the files, select the Run/Run command in the Discover module (lower menu bar). When the parameter block appears, go to the Parameters value-aid, change the Assembly/Mol Level to Assembly, and select ETH1_ALL. Make sure Computation Mode is set to Command File, toggle the PBC option on, and select Execute.

There is no need to worry about the other parameters, since you will copy the input file from the in vacuo run.

10. Modifying the .inp file

To modify the .inp file you leave Insight and enter the UNIX operating system level. Since the calculations for the gas phase and especially the aqueous phase take a long time, you will also run them in the standalone mode.

On the command line near the bottom of the screen, type quit on and press <Enter>. In the UNIX shell window, move (cd) to the directory in which you were running Insight (if you are not already there). Use any text editor to edit the eth0.inp file by adding the minimize command. For example, start editing the .inp file by entering: > vi eth0.inp

(You may use some other text editor if you prefer, but the instructions given here are specific for vi.)

Insert the minimize command between the calculate command and the initialize dynamics command. This is done by entering: > /initialize

to find that line by moving the cursor to it, followed by entering a capital O:

> O

to open a new line above it.

Then type the following lines, starting in the indicated columns: 123456789... minimize no cross no morse * for 200 steps using steepest descent * until maximum derivative is less than 1.0 minimize no cross no morse * for 1000 steps using conjugate gradient * until maximum derivative is less than 0.1

You also need to change the cutoff parameters:

Set cutdis to 7.8 and swtdis to 2.0. This can be done by moving the cursor to the appropriate number, typing R, typing the new value, and pressing the <Esc> key. To make the output file small, you can change the frequency at which the energy averages are printed out. To do this, open a new line in the initialize dynamics command, before the line * write history ... , and enter, starting in the indicated columns: 123456789... * averages every 100 steps Now depress the <Esc> key and write out this file by entering :wq. Create eth1_cell1.inp (the aqueous phase input file) by copying the eth0.inp file. Do this by entering: > cp eth0.inp eth1_cell1.inp

11. Running the Discover job

Before running the gas phase job, you must check that the .inp files are correct. Also check the first few lines of the eth1_cell1.car file for a line specifying that periodic boundary conditions are on (PBC=ON) and for a second line, starting with PBC, that gives the cell parameters. Next, check whether you have enough free disk space.

The space needed for a free energy run depends primarily on the number of atoms in the system, the number of quadrature points, the total number of steps of dynamics, and the frequency at which history files are written. For the parameters specified for this run, the gas phase calculation takes about 1 megabyte and the aqueous phase takes about 10 megabytes. (For an estimate of the size of history files with different parameters, refer to the documentation of the calculate relative command.) Allow an additional 2-3 megabytes to accommodate the .out, .tot, .wrp, .RS[1-n], and other regular Discover output files.

Finally, check whether you have enough CPU time available.

The gas phase calculation takes about 40 minutes on a Silicon Graphics Personal IRIS (4D20), and the aqueous phase takes about 3.5 days on the same machine.

When you are ready to start the Discover simulation, type: > discover eth0 std and answer the prompts by pressing <Enter>.

This starts the gas phase calculation.

When the gas phase job finishes, type: > tail -30 eth0.out to display the last 30 lines in the output file (Figure 2).

The free energy (dA(Lambda), column 6) and error (column 7) are tabulated for each Lambda (column 1), and totalled at the bottom (Figure 2).

Figure 2. Sample free energy summary table for the perturbation of methanol to ethane in a vacuum, gas system

SUMMARY OF RELATIVE FREE-ENERGY CALCULATION Integration of dA/dLambda vs. Lambda using FDTI 10.000 ps equilibration 10.000 ps sampling Sampled every 10 steps dLambda = 0.00500 Lambda range from 0.00000 to 1.00000 dA+/dLambda = -kT * ln <exp [-(H(Lambda+dLambda)-H(Lambda)) / kT]> / dLambda dA-/dLambda = +kT * ln <exp [-(H(Lambda-dLambda)-H(Lambda)) / kT]> / dLambda dA(Lambda) = [dA+/dLambda + dA-/dLambda] / 2.0 * weight LAMBDA dA+/dLambda dA-/dLambda error weight dA(Lambda) error sum(dA) (+/-) (kcal/mol) (+/-) (kcal/mol) ======= =========== =========== =========== =========== =========== =========== =========== 0.03377 33.47 33.72 0.57 0.0857 2.88 0.05 2.88 0.16940 15.58 15.45 0.13 0.1804 2.80 0.02 5.68 0.38069 14.00 13.88 0.11 0.2340 3.26 0.03 8.94 0.61931 12.08 11.97 0.11 0.2340 2.81 0.03 11.75 0.83060 12.07 11.95 0.14 0.1804 2.17 0.03 13.92 0.96623 14.75 14.61 0.23 0.0857 1.26 0.02 15.17 -------------------------- TOTAL 15.17 +/- 0.17

Now proceed to the aqueous phase calculation by typing: > discover eth1_cell1 std

As mentioned, this run may take 3-4 days.

When the simulation has completed, type: > tail -30 eth1_cell1.out to look at the last 30 lines of the .out file (Figure 3).

Figure 3 . Sample free energy summary table for the perturbation of methanol to ethane in a vacuum, solvated system

SUMMARY OF RELATIVE FREE-ENERGY CALCULATION Integration of dA/dLambda vs. Lambda using FDTI 10.000 ps equilibration 10.000 ps sampling Sampled every 10 steps. dLambda = 0.00500 Lambda range from 0.00000 to 1.00000 dA+/dLambda = -kT * ln <exp [-(H(Lambda+dLambda)-H(Lambda)) / kT]> / dLambda dA-/dLambda = kT * ln <exp [-(H(Lambda-dLambda)-H(Lambda)) / kT]> / dLambda dA(Lambda) = [dA+/dLambda + dA-/dLambda] / 2.0 * weight LAMBDA dA+/dLambda dA-/dLambda error weight dA(Lambda) error sum(dA) (+/-) (kcal/mol) (+/-) (kcal/mol) =========== =========== =========== =========== ========== =========== =========== =========== 0.03377 74.66 87.01 3.45 0.0857 6.92 0.30 6.92 0.16940 26.11 26.43 0.62 0.1804 4.74 0.11 11.66 0.38069 17.91 17.95 0.38 0.2340 4.20 0.09 15.86 0.61931 14.23 14.27 0.43 0.2340 3.33 0.10 19.19 0.83060 14.70 14.58 0.36 0.1804 2.64 0.07 21.83 0.96623 15.86 15.74 0.43 0.0857 1.35 0.04 23.19 -------------------------- TOTAL 23.19 +/- 0.70

12. Interpreting the results

To determine the solvation free energy difference for this system, go back to the thermodynamic cycle. Take the value for the solvated system (Figure 3) and subtract the value for the gas system (Figure 2). Compare the result with the experimental value.

The experimental value is 6.93 kcal mol^-1. In the example, the calculated value is 9.61-1.45 = 8.16 kcal mol^-1. The statistical error is 0.58 + 0.19 = 0.77 kcal mol^-1. Note the difference in the numbers between the dA+/dLambda column and the dA-/dLambda column. If the difference is large, then either the convergence is poor or the dLambda chosen is too large.

Note that the free energy is a statistical quantity and is thus highly dependent on the ensemble average. Because of this, the values you obtain in this lesson will be different from those in this example.

13. Checking the convergence

Convergence can be monitored by plotting the batch average of exp (-dH\xda kt) vs. time. This can be done by using the .tbl file with Insight.

Start Insight by entering insightII at the UNIX prompt, in the same directory in which you ran Discover. Select the Graph/Get command (from an icon). When the parameter block appears, select the File Name parameter and pick eth0.tbl in the Files value-aid. Select Time1 as the X Function, and Exponential1 as the Y Function. Select Execute when the Z Function parameter block is highlighted.

A plot showing the batch average of exp (-dH/kT) vs. time for the first lambda appears on the screen.

To obtain the convergence plot for the second lambda, follow the same procedure as above but select Time2 as the X Function, and Exponential2 as the Y Function.

The convergence plot should show a line fluctuating about a constant value. Any upward or downward drift indicates a convergence problem. You can also plot column 5 vs. column 2 in the .tot file. Another way to check convergence is by making the run longer and noting whether or not the results are stable.

14. Exiting the Insight program

On the command line, enter quit on and press <Enter>.

Lesson 3: Using pseudoatoms

Lesson 4: Consensus dynamics/minimization

Lesson 5: Minimizing the magainin peptide--standalone mode

Enter biosym_tutorial -i at the system prompt and answer the questions: If the Discover tutorial files are listed as not installed, enter discover at the first prompt. To get out of the installation menu, press <Enter>. Enter 3 to change to your ~/tutorial directory. In the same menu, enter 4 to obtain a system prompt. Now type discover and answer the prompts, as shown in the following box.

> discover Discover Molecular Modeling System, version 2.97 Note: Discover may be run without prompts by entering: discover file_name ff_name nice_number start_yn ncpus file_name = root name for the run ff_name = force field library file nice_number = background nice number start_yn = yes to start running in background ncpus = number of CPUs for concurrent execution e.g.: discover acenm /usr2/discover/libraries/cvff.bin 19 y 2 or discover acenm std 19 n 4 In order to use DISCOVER you will need three files with the same root name and the following standard extensions. suffix .inp A file of DISCOVER commands to control the simulation. .car A file with cartesian coordinates, atom and residue names. AND, one of these two files: .mdf or A formatted (.MDF) or unformatted (.RST) file of .rst connectivity information corresponding to the system to be modeled. A .MDF file can be output from INSIGHT; a .RST is an output file of DISCOVER. >> ENTER the file name PREFIX (<RETURN> = QUIT) >> magainin_helix Which potential function do you wish to use? 1) library $BIOSYM/irix51R4/biosym_lib/cvff.bin 2) library $BIOSYM/irix51R4/biosym_lib/amber.bin 3) library $BIOSYM/irix51R4/biosym_lib/cff91.bin 4) library $BIOSYM/irix51R4/biosym_lib/pcff.bin 5) other >> ENTER 1, 2, 3 or 4, (<RETURN> = 1) 1 >> ENTER the nice number for running discover as a background process. The default is 19 >> 19 >> Do you wish to start discover now? The default is YES >> yes Running magainin_helix.csh as a background process... The output from discover.exe will go to magainin_helix.out. The error from the operating system will go to magainin_helix.log. [1] 2523

Notice that your job has been assigned a number, [1] 2523, where [1] is the relative number of the job you submitted, and 2523 is a job identification number. You can now enter 5 to exit the biosym_tutorial.

2. Checking the progress of the job

Enter: > ps -e | grep 2523

(The number assigned to your job will probably not be 2523; use the actual number assigned.) If it is still in progress, the job identification number and time appears on the screen. If only the system prompt appears, the job has completed.

A .log and .out file are created as soon as your Discover job begins to process.

3. Analyzing the results

When your job completes, print the standard output file (.out). Identify: Which residues have the most internal strain? Which energy component decreased the most during minimization? What was the initial tethering energy? What was it at the end? What was the rms deviation during the initial template forcing? Which backbone torsions changed the most during minimization? Which side chain torsions changed the most during minimization? How efficient was the minimizer in converging? Did it converge?

The tethering energy is included in the sum listed in the output as "total function value minimized" and can be obtained as the difference between this sum and the sum listed under "total energy".

Lesson 6: Conformational search by high-temperature dynamics--standalone mode

1. Starting the tutorial

Enter biosym_tutorial -i and answer the prompts. If the Discover tutorial files are listed as not installed, enter discover at the first prompt. To get out of the installation menu, press <Enter>. Enter 3 to change to your ~/tutorial directory. In the same menu, enter 4 to obtain a system prompt. Now type discover and answer the prompts, as shown below. > discover Discover Molecular Modeling System, version 2.97 Note: Discover may be run without prompts by entering: discover file_name ff_name nice_number start_yn file_name = root name for the run ff_name = force field library file nice_number = background nice number start_yn = yes to start running in background ncpus = number of CPUs for concurrent execution e.g.: discover acenm /usr2/discover/libraries/cvff.bin 19 y 2 or discover acenm std 19 n 4 In order to use DISCOVER you will need three files with the same root name and the following standard extensions. suffix .inp A file of DISCOVER commands to control the simulation. .car A file with cartesian coordinates, atom and residue names. AND, one of these two files: .mdf or A formatted (.MDF) or unformatted (.RST) file of .rst connectivity information corresponding to the system to be modeled. A .MDF file can be output from INSIGHT; a .RST is an output file of DISCOVER. >> ENTER the file name PREFIX (<RETURN> = QUIT) >> magainin_beta Which potential function do you wish to use? 1) library $BIOSYM/irix51R4/biosym_lib/cvff.bin 2) library $BIOSYM/irix51R4/biosym_lib/amber.bin 3) library $BIOSYM/irix51R4/biosym_lib/cff91.bin 4) library $BIOSYM/irix51R4/biosym_lib/pcff.bin 5) other >> ENTER 1, 2, 3 or 4, (<RETURN> = 1) 1 >> ENTER the nice number for running discover as a background process. The default is 19 >> 19 >> Do you wish to start discover now? The default is YES >> yes Running magainin_beta.csh as a background process... The output from discover.exe will go to magainin_beta.out. The error from the operating system will go to magainin_beta.log. [1] 2524

Notice that your job has been assigned a number, [1] 2524, where [1] is the relative number of the job you submitted, and 2524 is a job identification number. Enter 5 to exit the biosym_tutorial script.

2. Displaying the minimized structure with Insight

Once the run is finished, the minimized model can be read in and displayed using the Insight molecular modeling program. The file containing the minimized coordinate is named magainin_beta.cor. The trajectory explored during the simulation may be displayed in Insight using the Animate command. The trajectory data are contained in the file magainin_beta.his.

To exit Insight, enter quit on the command line near the bottom of the screen and press <Enter> twice.