Skip to main content
Center for Biomolecular NMR Spectroscopy
HomeUser GuidelinesSpectrometersSchedulingFeesUNIX Software InformationEducation and USER BASELinksContact Us
Aria Tutorial: Running a Calculation with PIPP data

Aria is started by typing aria at the UNIX shell promt. Instead of being a command name, aria is actually an alias that sends a script written in the Python language (provided by ARIA) to the Python interpreter. This script looks for either new.html or run.cns files in the directory from which ARIA is called, parses the files, and executes the appropriate commands. At first, new.html file should be created and ARIA started in the same directory. New.html contains the locations of all the original data files that ARIA will use. When started, the program reads in all the given data, converts it to a format that ARIA can use, creates a directory called run1 and puts all the converted data there. Run1run.cns, which defines all the parameters for the calculation. This file needs to be edited to suit your needs (specify which computers you want to use, specify disulfide bonds, etc.). Once run.cns is edited, ARIA should once again be started (in run1 this time), and this will start the calculation. Both new.html and run.cns can be edited either manually (in a text editor) or through the user-friendly web-based interface that is provided with ARIA.

Create new.html

new.html is a file used by Aria to locate the files it needs for calculations (chemical shifts file, .PCK files, etc.). Go to the webpage located at file:/u/progs/aria/aria1.0/html/aria.html and click on the option that says "start a new project with PIPP (uncalibrated data)." This will bring up a form that creates the file new.html.

1. Enter the "ARIA path" as /u/progs/aria/aria1.0

2. Enter your project directory (this should be the directory where you intend to have the calculation.

3. The "spectrum name" is completely arbitrary and will only alter the name of the directory and the files (e.g. "spectrum name" 3dc will produce files like 3dc.tbl) into which the data is placed.

4. Enter the absolute path of the .PCK file you intend to use.

NOTE: Look at your .PCK file. the very first peak in the list needs to be assigned. If necessary, copy a peak from somewhere else in the list and place it as the first one. If the first one is not assigned, ARIA will give you an error trying to read the file.

NOTE: The data offset (set in stapp.par is not read in by ARIA, so it has to be added to the coordinates manually using an awk script (see example).

5. Enter the absolute path for the chemical shifts file (such as junk.shifts)

6. In the pull-down menus, specify which atoms are located in which column of the coordinates of the PCK file (e.g. for 3dc, the carbon (or heteronucleus) is usually in dimension 3). Usually the defaults are OK for 3D experiments.

NOTE: a "heteronucleus" is a non-hydrogen atom, so if the heteronucleus1 is a carbon atom, proton1 would be the hydrogen atom attached to it.

7. Specify the PPMD (the error range) for each atom. The error range is usually specified in stapp.par file, and even though ARIA does read the file in, it does not read any of the parameters in it, except for the sequence.

8. Repeat for as many spectra as is necessary.

9. Specify the sequence file (such as stapp.par ) and other optional files, such as dihedran angle restraints and dipolar couplings.

The sequence file can be a .pdb file, a .par file, or any other format, as long as it contains the protein sequence (one-letter code with line breaks, e.g. ACLG \ YK). ARIA does not read any parameters from the file other than its sequence.

10. Click on the "Save updated parameters" button to save the file. It will display something like /dir/dir/start_project.cgi . Make sure the path that it is showing is to your project directory and instead of start_project.cgi save it as new.html .

Run ARIA

Go to the directory of your project and type aria at the command prompt. This should create a directory called run1 and convert and copy all of the converted data files into that directory to be later used in actual calculations. The file run.cns is also generated and put into run1 .

NOTE: At this point in time, ARIA cannot use 2D or 4D PIPP spectra.

Edit run.cns

The newly edited run.cns contains all the parameters ARIA uses in the actually calculation. Once again, go to the website file:/u/progs/aria/aria1.0/html/aria.html and either enter manually or browse to the location of your run.cns file and click on "Edit file." Although run.cns has numerous options, there are several key points that need to be edited:

1. Make sure the conversion to DIANA nomenclature is set to "false" unless you are using XEASY data. Since ARIA is optimized for XEASY, this option is defaulted as "true."

2. Enter the number and location of disulfide bonds (if any). Upon the initial conversion and calibration, ARIA takes out diagonal peaks and disulfide bonds, so they have to be input manually.

3. Specify which machines you want ARIA to run on. Specifying more than one machine involves using the rsh command. The first one (on local machine) should look something like:

{===>} queue_1="/bin/csh /instinct1/lerap/aria/run1/";

{===>} cns_exe_1="/u/progs/cns_solve_1.0/sgi-r8k10k-irix-6.5/bin/cns_solve";

{===>} cpunumber_1=1;

and the rest should look like

{===>} queue_2="rsh inept.v24.uthscsa.edu /bin/csh /instinct1/lerap/aria/run1/";

{===>} cns_exe_2="/u/progs/cns_solve_1.0/sgi-r8k10k-irix-6.5/bin/cns_solve";

{===>} cpunumber_2=1;

For other machines, you will need to change the machine name in the command (i.e. inert.v24.uthscsa.edu etc.). . Also, make sure that the rsh command contains the path to your project directory. The "cns_exe" field contains the path to the CNS executable file, and the "cpunumber" field is used to input how many processors are on that particular machine (at this time all machines on the v24 network have only one).

4. Save the file. It will once again display /dir/proj_dir/cns_frm2inp_0.3.cgi. Edit the path appropriately (the location of your run1 directory) and make sure cns_frm2inp_0.3.cgi is replaced with run.cns.

Run ARIA

Go to the directory where run.cns is located and type aria at the UNIX shell prompt. This will start the actual calculation. ARIA can be stopped at any point and when restarted, will start again at the point where it was stopped. Do NOT log out when ARIA is running. Even set in background, ARIA will quit if you log out.

Analyzing results

In each of eight iterations, ARIA makes twenty structures through simulated annealing. Each structure is a .pdb file, and the names of all the structures are listed in the file file.nam, which can be opened with Molmol and will load all the structures listed within it. Each iteration also has a .tbl file for each spectrum; these are lists of restraints used in the calculation during the current iteration. The .list files contain all the restraints, including violations. Iteration 8 contains the data for the final energies, geometries, etc. analyses and water-refined structures. In the last iteration, all the restraints are converted to Molmol format (.lol and .upl), violations are printed out, and a Ramachandran plot is created. During the course of the entire calculation, all the output of CNS scripts is written to an appropriate .out file (e.g. output of print_noe.inp is written to print_noe.out ). Since the cns output files are long and messy, there is a great script that can be used to filter out the more important information. This script will be included with the next release of ARIA. For now, copy the ariaoverview.pl file to your run1 directory

cp /u/people/lerap/ariaoverview.pl .

and execute it:

/usr/bin/perl ariaoverview.pl

It will read in the .out files from all the iterations, create several plots, and print out violations, energies, geometries, and an overall overview of the calculation. This information is put into overview.* files.



Friday, 03-Aug-2001 16:34:48 CDT [an error occurred while processing this directive]