IPA (1) User Commands IPA (1) NAME ipa - find a set of alignment parameter values that makes the example costs of input protein multiple alignments close to optimal SYNOPSIS ipa [OPTION]... FILES DESCRIPTION Given a set of protein sequence multiple alignments, write to the standard output a set of alignment parameter values (gap-open cost, gap-extension cost, and substitution cost matrix if one is not provided) such that under an affine alignment cost function with these parameter values, the costs of examples from input multiple alignments close to optimal. FILES Use FASTA formatted text files FILES to find alignment parameter values. If any file of FILES doesn't contain core block(s), the iteration process is skipped. Following OPTIONs are not madatory and the order of OPTIONs is not important. -h, --help Display help message to the standard error. -i, --integer[=SCALE] Report the alignment parameter values in integer interval [0, SCALE] instead of [0, 100]. The default value for SCALE is 100. -n, --no-integer Report the alignment parameter values as they are. This takes precedence over --integer=SCALE. It's not recommended to use this option unless one knows what she is doing. -o, --output[=FILE] Write the alignment parameter values to file FILE. The default value for FILE is the standard output. If this option is not used, the alignment parameter values are reported to the standard output. -p, --precision[=P] Use real value P as a precision such that any error (including their bounds) is fit into the precision. P should be in the real interval [10^(-5), 10^(-3)] and should be a power of 10. The default value for P is 10^(-3). -a, --absolute Use the absolute error criterion. This is the default error criterion. If provided with --relative[=R] option, --absolute takes the precedence over --relative[=R]. -r, --relative[=R] Use the relative error criterion. If real value R is provided, the smallest relative error is not searched but the error is fixed at R. R should be in real interval [P, 1] (read --precision[=P] option). If provided with --absolute option, --relative[=R] is ignored. -u, --upper-bound[=U] Use real value U as an upper-bound on the smallest relative error if the search occurs. The default value for U is 1. U should be in real interval [L, 1] (read --lower-bound[=L] option). If provided with --absolute or --relative=R option, --upper-bound[=U] option is ignored. -l, --lower-bound[=L] Use real value L as a lower-bound on the smallest relative error when the search occurs. The default value for L is 10^(-3). L should be in real interval [P, U] (read --precision[=P] and --upper-bound[=U] options). If provided with --absolute or --relative=R options, --lower-bound[=L] is ignored. -c, --common-substitution[=VALUE] Use commonly used fixed substitution cost matrix specified by VALUE, which can be one of the following: blosum22, blosum35, blosum45, blosum50, blosum50_2, blosum55, blosum60, blosum62, blosum62_3, blosum62_4, blosum65, blosum70, blosum75, blosum80, blosum80_3, blosum85, blosum90, blosum95, blosum100, blosum100_3, pam120, pam160, and pam250. The default value for VALUE is blosum62. -f, --fixed-substitution=COST Use the alphabet and the substitution cost matrix specified in BLOSUM formatted text file COST. -s, --statistics=FILE Write various statistics collected during the alignment parameter value search to file FILE. -g, --plot=FILE Write a GNU Plot formatted text describing the search of the final alignment parameter values to file FILE only if the substitution costs are fixed. Otherwise this option is ignored. Once FILE is produced, one can use GNU Plot to see the graph. -t, --verbose[=FILE] Write a detailed progress report to file FILE. The default value for FILE is the standard error. -v, --version Write the version of the program to the standard error. IPA 3 May 2007