Introduction.

CCOMP is a simple command-line program for comparing ligand/receptor complexes. The program can be used for calculating pairwise all-atom RMSD of slightly different protein structures, taking care of missing atoms, sequence inconsistencies, and other problems often encountered in PDB files. CCOMP can read two files in PDB format, including both a receptor and a ligand, compute a pairwise sequence alignment of the receptor molecules, generate alpha-carbon superposition of the receptor structures according to the generated alignment, and calculate individual deviations per residuum. Additionally, several statistical measures are provided in the input file.


Compilation.

Source code of the CCOMP program is available in /src directory. CCOMP can be compiled using standard GNU C compiler:

cc -O2 ccomp.c -lm -o ccomp

Pre-compiled binary versions of CCOMP are provided within program's archive in /bin subdirectory.


Running CCOMP.

CCOMP can be run from a command line in a following way:

./ccomp <structure1.pdb> <structure2.pdb> > <output file>

where structure1.pdb and structure2.pdb are files including structures of the analyzed complexes in PDB format. The programs expects the receptor molecules to be specified as first molecules in both input files. Several common errors and inconsistencies often present in PDB files, such as multiple occupancy sites, chain breaks, missing residues, sequence mismatches, are automatically recognized and fixed by CCOMP.

For example, the following command:

./ccomp 1db1.pdb 1ie9.pdb > output.txt

will generate a text output file output.txt.

By default, only side chains exhibiting significant deviations (z-score > 1) are output by the program. Optional "-all" flag can be specified as a third command line parameter in order to display differences between all side chains.

Additionally, CCOMP generates a structural superposition file complex_superimposed.pdb that can be read and displayed by molecular graphics program in order to provide visual feedback on analyzed complexes.


CCOMP output file.

The output file is divided into several sections:

* general structure information:

- names of input files
- receptor chain lengths
- amino acid sequences of receptor proteins
- experimental structure resolutions of the receptor molecules

* sequence comparison results of both receptor sequences:

- sequence alignment
- alignment score
- alignment length

* global structure comparison results:

- global root mean square deviations (RMSD) calculated over alpha carbon atoms and all atoms
- average center of mass (CM) and equivalent distance (DISTANCE) deviations calculated for whole amino acids and for side chain atoms

* detailed analysis of differences between individual amino acids in both complexes:

- deviations and normalized deviations of coordinates (L), distances (D) and centers of masses (C) of individual amino acids and their side chains. The normalized values (z-score, Z) are calculated by dividing observed distribution of deviation measures by standard deviation of expected distribution.
- analysis of Chi-angle deviations between individual amino acids
 
* analysis of position deviations of ligand molecules in both complexes

* analysis of receptor - ligand contacts in between heavy atoms in both analyzed complexes.