First, a big acknowledgement to Steve Searle (European Bioinformatics Institute)
for getting STAMP to run (finally) under OSF and 64 bit machines generally.
Also thanks to Andrew Torda
(Australian National University, Canberra), Dave Schuller (University of California-Irvine),
Mike Tennant (SmithKline Beecham Pharmaceuticals, Harlow, UK),
Asim Siddiqui (LMB, Oxford) G.P.S. Raghava (LMB, Oxford), and James Cuff (EBI)
for their help and various painstaking trawls through my spaghetti code.
Apart from bugs, etc., the noticable changes are:
1. STAMP now reads compressed PDB and DSSP files. It will also look for files that are stored in a brookhaven style directory structure (e.g. distr/mb/pdb4mbn.ent).
2. Output is now flushed (fflush) during scanning. Purely a cosmetic thing for those who want up to the minute output when the program is running.
3. AVESTRUC now has an option to calculate an average for all aligned positions. It also now outputs values to the temperature factor fields in the PDB output to denote those averaged positions corresponding to structurally equivalent regions (blue in RASMOL colour by temperature) and those equivalenced fortuitously (red). It will also highlight positions showing identical or conserved residue character.
4. PDBSEQ has some new options, including the ability to output separate files for each domain, and now outputs a sensible description of the protein by considering the TITLE, COMPND and SOURCE entries in each PDB file. Note that the default format is now FASTA.
5. DSTAMP has been changed drammatically, and is now (I think) much more useful. The input files for ALSCRIPT are now much prettier, including cylinders/arrows for helices/strands and colouring/fonting according to residue property conservation within the sequence alignment. It can also be used on alignments not derived using STAMP (i.e. from GCG, AMPS etc.).
6. STAMP now appears to run smoothly under OSF. Once again thanks to Steve Searle. Versions have also been compiled and tested on IRIX, Solaris and Linux.
7. CLUS2BLC and SMSF2BLC have been replaced by a general alignment conversion program written in PERL (ACONVERT). More details are given below. Note that this is a general alignment conversion utility that might be useful in other contexts apart from STAMP.
8. The significance of sequence identity following structural alignment is now estimated according to Murzin (1993), JMB, 230, 689-694.
9. Three new programs have been added to the package (see specific instructions below):
MERGETRANS allows one to combine transformations from a variety of different sources (i.e. ALIGNFIT and STAMP). It either uses a user-specified identifier to link the two files (i.e. one found in both files) or the first common identifier if none is specified.
MERGESTAMP. Like MERGETRANS this program permits one to merge various kinds of STAMP data. However, it considers more than merely the transformations, and attempts to combine the alignments as well. It can be used in exactly the same way as MERGETRANS (i.e. to combine files that contain only transformations), but will also attempt to merge alignments in the file, if they are present. The alignments must be in BLOCK format (see the depths of the manual for details, and for how to convert things like Clustal or MSF into BLOCK format). MERGESTAMP can combine files that do not contain transformations as well (i.e. those that contain only alignments), and can thus be used for sequence data handling as well.
EXTRANS allows one to select and extract particular domains from a transformation file.