2. Quick Start
2.1. Working directory and control file "run.prm"
MCCE is a command line program. MCCE expects a control file named "run.prm" and will write out results as well as temporary files in the directory where the mcce command is invoked. This directory is called "working directory". Usually the "working directory" is set up to hold a pdb file and a "run.prm" file to start the calculation. We recommend one working directory for one structure.
The control file "run.prm" gives instructions to the mcce program. This file must be named as "run.prm" and placed in the working directory. Two rules apply to this file: 1) Any line starting with "#" is a comment line and will be ignored; 2) If a line is not a comment line, MCCE interprets the line as a key-value pair in which the string in parentheses is the key and the first string of this line is the value. Here are the sections of the often customized lines of the control file "run.prm":
Most modified entries
Input and Output:
f step 1: pre-run, pdb->mcce pdb (DO_PREMCCE)
f step 2: make rotatmers (DO_ROTAMERS)
f step 3: do energy calculations (DO_ENERGY)
f step 4: monte carlo sampling (DO_MONTE)
Less modified entries
8.0 Protein dielectric constant for DelPhi (EPSILON_PROT)
/usr/local/mcce/name.txt MCCE renaming rule. (RENAME_RULES)
ph "ph" for pH titration, "eh" for eh titration (TITR_TYPE)
0.0 Initial pH (TITR_PH0)
1.0 pH interval (TITR_PHD)
0.0 Initial Eh (TITR_EH0)
1.0 pH interval (TITR_PHD)
0.0 Initial Eh (TITR_EH0)
30.0 Eh interval in mV (TITR_EHD)
15 Number of titration points (TITR_STEPS)
In the most often modified section, the input structure must be specified by the line "(INPDB)". Change the name prot.pdb to the name of your file. Although a pdb file in working directory is recommended, the file can be in a distant directory with the full path in the name.
There are 4 steps in a MCCE run. You can run run them all at once or stop between steps. Each step is dependent on output of the previous ones written into the working directory (see Figure 1, file flow of mcce, which shows needed input and output files for each step). To run a step of MCCE, mark its flag as "t".
In the less often modified section, you may specify the dielectric constant of the protein in line "(EPSILON_PROT)". Currently, dielectric constants 8 and 4 are supported, but it is expandable. To use a different dielectric constant, one needs to find reference reaction field energy for residues to transfer from the aqueous solvent to the protein with that dielectric constant and put them into the parameter files. Please ask us and we will add the instruction to the manual. Normally dielectric constant 8 is good for small soluable proteins (< 200 residues) and dielectric constant 4 is good for big transmembrane proteins (>200 residues). The solvent is always assigned a dielectric constant of 80.
The line "(EXTRA)" points to a file with offsets for each residue type calibrated by comparing about 630 calculated pKas and experimental pKas. The origin of the system error is possibly from inaccurate solution pKa of residues, different solvent entropy effect and different polarization of atomic charges between the reference system water and protein material. Leave this line empty if you don't want to use this correction. This file is a sumplemental parameter file to the parameter directory. This file follows the same format as files in the parameter directory and entries in this file have higher priority than those in the paramerter directory. This provides means to override the entries in the mcce parameter directory, providing a convenient way to test a new parameter file before depositing it to the permanent parameter directory.
The file specified in "(RENAME RULES)" allows you to rename, cut and combine residues to defined cofactores. It is sometimes necessary to split a cofactor into several ionizable groups or combine ligand residues with cofactor to be one group. For example, you may want to treat a His ligand of a heme differently than regular His or treat the propionic acids on a heme as separate titratable groups. This can be achieved by renaming the atom name, residue name and sequence number of these residues. A sample rename file "name.txt" is in the MCCE distribution directory. When a new cofactor type is created by the recombination of residues, a parameter file similar to those in the parameter directory should be created.
The Monte Carlo Sampling can do either a pH titration or a redox potential (Eh) titration. The above sample lines configure a pH titration from pH=0 to pH=14 with a step of 1 pH unit.
The line "(MCCE_HOME)" defines the location of mcce program and the parameter directories. If you have more than one districbution of mcce programs, you may want to switch between the programs and the parameter directories. For now, the only use of this line is to compose the full path of the parameter directory. For example, if this line points to "/usr/local/mcce" and "(EPSILON_PROT)" gives dielectric constant 4, then the parameter directory is "/usr/local/mcce/param04".
There are 3 sample "run.prm" files included in the MCCE directory. They are different in the detail level of making rotamers. "run.prm.quick" is suitable for a quick and dirty pKa calculation which is done with minimal number of rotamers. "run.prm.default" provides moderate number of rotamers and predict more accurate pKa values than "run.prm.quick". "run.prm.full" will produce extensive rotamers, predict accurate pKas, and is good for exploring side chain motions. Here is a comparison of computing time and calculated pKas of Hen lysozyme (pdb 4lzt):
Table 1. Comparison of runs on lysozyme with the three run.prm presets. When comparing these numbers with experimental pKa values, bigger errors are often seen for the binding sites and functionally important residues marked by *. These residues often show abnormal pKa values which are hard to calculation. An error about 1 pH unit on these sites is often seen.
|Presets of run.prm
|Number of conformers
|Step 1 - check protein
|Step 2 - make rotamers
|Step 3 - generate lookup table
|Step 4 - run Monte Carlo
|Overall RMSD of error
<Prev | Contents | Next >