27 October 2009           Grishin Lab, compass_search directory

The compass_multisearch directory contains the following subjects:

README:                  This README file.
solaris_intel/ :         codes mk_compass_db and compass_db1Xdb2_search_2_3 compiled in Solaris (Intel) system  
linux/ :           	 codes mk_compass_db and compass_db1Xdb2_search_2_3 compiled in LINUX (glibc version 3.3)
irix/ : 		 codes mk_compass_db and compass_db1Xdb2_search_2_3 compiled in IRIX 6.5

COMPASS is the program for the comparison of two multiple protein alignments

(Sadreyev, R. I. & Grishin, N. V. (2003) COMPASS: a tool for comparison 
of multiple protein alignments with assessment of statistical significance. 
J Mol Biol 326, 317-336).

Codes mk_compass_db and compass_db1Xdb2_search allow running multiple searches,
with a set of alignments as queries, against another set of alignments. 


mk_compass_db (version 1.3):
     Reads alignments specified in a file list,
     creates their numerical profiles (to be used by COMPASS),
     and writes the profiles into a single output file.
     The total filtered length of the database (excluding positions with high gap content)
     is written into a separate file.

You can get the following HELP by starting the code with no arguments 
(simply typing mk_compass_db in the command line):  

Required arguments:

  -i    Input file with list of alignment files to read
        (if no file is specified, all ./*.aln files are used).
        Alignment format: ClustalW, STOCKHOLM, or simple alignment format
  -o    Output file with profile batch (Neffs for each position and PSSM)

Optional arguments:

  -q    Path to the matrix of residue pair frequencies (q_ij)
        Default = BLOSUM62
  -g    Threshold of gap content to disregard 'gapped' columns (0.0 to 1.0)
        Default = 0.5
  -l    Lower bound of ordering number of files in the subset to be processed from the whole list
        Default = 1
  -h    Higher bound of ordering number of files in the subset to be processed from the whole list
        Default = total number of files in the list




compass_db1Xdb2_search (version 3.1):
Runs comparison of two sets of numerical profiles against each other,
with output formatted as separate search results for each profile of the 1st set as a query

You can get the following HELP by starting the code with no arguments: 

Arguments:

  -i   Input numerical database of query alignments[File in]
        (precomputed from a set of alignments using program make_compass_db)

  -j   Input numerical database of subject alignments [File in]
        (precomputed from a set of alignments using program make_compass_db)

Optional arguments:

  -o    Output file (default = STDOUT) with COMPASS result:
        a header
        and resulting alignment (top sequences shown in both input alignments):
        pluses (+) denote matches with positive scores;
        Capital letters and dashes (-) denote the profile positions
        that were used for alignment construction.
        Lower-case letters and dots (.) denote the profile positions with high gap content
        that were disregarded in the process of alignment construction
        (they may be aligned with gaps (~) that are not scored).
        Equal signs (=) denote the gaps introduced in profiles.
  -s    Path to the residue substitution matrix (s_ij)
        Default = BLOSUM62
  -q    Path to the matrix of residue pair frequencies (q_ij)
        Default = BLOSUM62
  -g    Threshold of gap content to disregard 'gapped' columns (0.0 to 1.0)
        Default = 0.5
  -t    Threshold of gap content for marking 'moderately gapped' regions,
        where penalty for the 1st gap opening is waived (0.0 to 1.0)
        Default = 1.0 (no waiving of gap opening penalty)
  -e    Maximal Evalue ro display
        Default = 1e-3
  -v    Number of detected database entries to show for each query (integer)
        Default = 100
  -O    Penalty for gap opening (integer)
        Default = 10
  -E    Penalty for gap extension (integer)
        Default = 1
  -L    Ungapped lambda for a standard sequence-sequence scoring system
        used in the score rescaling for profile-profile comparison
        Default = 0.3176 (ungapped lambda for BLOSUM62)
  -x    Expected value of individual positional scores for a standard sequence-sequence
        scoring system; used in the score rescaling for profile-profile comparison
        Default = -0.5209 (expected value for BLOSUM62)
  -z    Database length (including only columns with gap content lower than the threshold (-g))
        used for Evalue calculation
        Default = length of alignment 2
  -b    Length of alignment chunks to print out in lines
        Default = 60
  -n    Number of sequences to print out in each of the aligned alignments
        Default = 1





For further information please contact:
Ruslan Sadreyev, sadreyev@chop.swmed.edu
Nick Grishin, grishin@chop.swmed.edu