Every entry in a STAMP input file is called a `domain'. This word
is a bit of a misnomer, since these things needn't be single domains
(though it is usually best to do structure comparisons at the domain level).
The problem of defining domains such that a wide variety of
possibilities may be used (e.g. all the coordinates in a PDB file,
one chain, bits of one chain, two chains, one chain and bits of
another, etc) is solved by defining a domain by: 1) a file, 2) an
identifier, and 3) a list of `objects', from the file, to be
included in the domain. An object is defined as a run of
coordinates, and a domain may contain more than one object.
Domains are stored in STAMP in files which may contain one or
more of such domain definitions.
The format of these files must be as follows:
<file name> <identifier label> { <objects> }
or,
<file name> <identifier label> { <objects> [RETURN] R11 R12 R13 V1 R21 R22 R23 V2 R31 R32 R33 V3 }
<file name> is the full name (including path) of the PDB file in
which the coordinate information is to be found. If you don't know the
precise location of the file, then just call it UNK or something (i.e.
not a blank), and the programs should be able to find the appropriate PDB
file using the identifier (if one can be found on your system), e.g.
/usr/people/jack/pdb4mbn.ent
<identifier label> is a short name to be used by the program.
eg. 4mbn1
If secondary structures are to be found by the program, then the first four letters of the identifier label should be the PDB code should correspond to the prefix used in your PDB/DSSP naming system. There should be no duplication of these, to allow for self comparison. It should contain the brookhaven four letter code first and anything else afterwards.
<objects> are coordinate descriptions, and may be one of three types:
1. ALLall 's from the file.
2. CHAIN Xonly 's labeled as chain X.
3. <chain1> <number1> <insert1> to <chain2> <number2> <insert2> e.g. B 20 _ to B 67 Ponly 's between (and including) the two full brookhaven
/data/newpdb/pdb/pdb1ton.ent 1ton { ALL 0.9876 0.34 0.543 19.23 1.0 2.34 0.98473332 1.0 0.023 0.94 4.345 20.0 } /data/newpdb/pdb/pdb2kai.ent 2kai_Kallikrien { CHAIN X CHAIN Y } /data/newpdb/pdb/pdb3sgb.ent 3sgbe_SGprotease { E 20 _ to E 160 P 1.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 1.0 0.0 }
Note the spaces. There must be spaces separating each keyword or datum to be read, even between the braces. For example:
/data/newpdb/pdb/pdb3sgb.ent 3sgb_protease{E 20 _ to E 160P}
would not be allowed.
In the second domain (Kallikrien) the transformation will be set equal
to the identity matrix with a translation of zero, since none has been
supplied.
The domains must be listed at the start of a file (ie. nothing must
come before them in a file), but anything may come afterwards,
provided that it contains no braces (ie. { or }) unless they are on lines
containing `%` in the first column.
It is possible to reverse the direction of an object in a domain
description. For example, if one has two objects,
one can reverse the direction of one or more of these by placing the
word "REVERSE" in front of the object, e.g.:
/data/newpdb/pdb/pdb4mbn.ent { REVERSE _ 1 _ to _ 20 _ _ 21 _ to _ 120 _ }