PROBCONS ~~~~~~~~ Probabilistic consistency-based multiple sequence alignment ----------------------------------------------------------------- PROBCONS is a novel tool for generating multiple alignments of protein sequences. Using a combination of probabilistic modeling and consistency-based alignment techniques, PROBCONS has achieved the highest accuracy of all alignment methods to date. PROBCONS was developed by Chuong B. Do in collaboration with Michael Brudno in the research group of Serafim Batzoglou, Department of Computer Science, Stanford University. For more information on the algorithms, please see Do, C.B., Brudno, M., and Batzoglou, S. (2004) PROBCONS: Probabilistic Consistency-based Multiple Alignment of Amino Acid Sequences. 12th International Conference on Intelligent Systems for Molecular Biology. In press. and Do, C.B., Brudno, M., and Batzoglou, S. (2004) PROBCONS: Probabilistic Consistency-based Multiple Alignment of Amino Acid Sequences. The 19th National Conference on Artificial Intelligence (AAAI-04). In press. ----------------------------------------------------------------- PROBCONS has been made freely available as PUBLIC DOMAIN software and hence is not subject to copyright in the United States. This system and/or any portion of the source code may be used, modified, or redistributed without restrictions. PROBCONS is distributed WITHOUT WARRANTY, express or implied. The authors accept NO LEGAL LIABILITY OR RESPONSIBILITY for loss due to reliance on the program. ----------------------------------------------------------------- Version History 1.0, 3/23/2004 (Chuong Do) -- initial release 1.01, 3/25/2004 (Chuong Do) -- fixed error in training procedure -- retrained default parameters for 1 and 2 pairs of insert states 1.02, 4/17/2004 (Chuong Do) -- replaced LOG_ADD and EXP routines -- added support for reading MSF format files -- added two extra utilities for scoring PROBCONS alignments (for benchmarking purposes) -- added the "compare" program for scoring alignments according to a reference alignment with respect to sum-of-pairs and column scores -- added the "fixref" program for adjusting PREFAB alignments to contain all letters of the input sequences; basically the main program for PROBCONS "hacked" to get the job done 1.03, 5/3/2004 (Chuong Do) -- added option to do all-pairs pairwise alignments instead of constructing a full multiple alignment -- added support for reading DIALIGN style files -- enabled support for using BAliBASE annotations for scoring BAliBASE alignments -- several minor bug fixes thanks to Bob Edgar -- added "project" program to project multiple alignment to pairwise alignments 1.04, 5/9/2004 (Chuong Do) -- switched over to default of one-insert state pair -- retrained default parameters -- added annotation scores -- small changes to model topology to make end gaps symmetrical -- added makegnuplot utility to plot annotation scores 1.05, 5/26/2004 (Chuong Do) -- added cutoff filtering for posterior scores -- made small corrections to recurrences for computing alignments -- added CLUSTALW output support 1.06, 7/13/2004 (Chuong Do) -- ProbCons is now PUBLIC DOMAIN software. 1.07, 8/30/2004 (Chuong Do) -- Fixed CLUSTALW output for sequence names (thanks to John Calley for pointing this out) 1.08, 8/31/2004 (Chuong Do) -- Added option for alignment order output (-a). 1.09, 9/1/2004 (Chuong Do) -- PROBCONS now allows input files with existing gaps -- these are automatically stripped before alignment. 1.10, 3/16/2005 (Chuong Do) -- Reduced memory consumption by -- not storing posterior matrix transposes -- restricting consistency-derived posterior matrices to original posterior matrix 1.11, 12/1/2006 (Chuong Do) -- Fixed bug with conflict between --annot and -a options 1.12, 4/1/2006 (Chuong Do) -- Fixed GCC 4.3 compatibility problems (thanks to Martin Michlmayr and Charles Plessy)