WHAT IF knows several ways of building, or changing structures. Some of them
are found in the SOUP menu, some in the DGLOOP menu, and some in the BUILD
menu. Building is not possible for residues that are not present in the
topology file.
This chapter describes the BUILD menu. This menu allows for crude building
from scratch;building by homology is done in the PIRPSQ menu.
This menu accepts amino acids in one- and in three-letter code.
Nucleic acids have to be provided in four-letter code (See the topology
file if you want to see these codes).
WHAT IF can only operate on residues that are present in its topology file.
This file can be found in the DBDATA directory, it is called TOPOLOGY.FIL.
If you do not know where
this directory is located, ask your local WHAT IF manager. The topology
file holds standard coordinates for many residues. Upon building WHAT IF
will use these standard coordinates, but it will rotate and
translate the molecule till it more or less fits where you want to build it.
The format of the topology file is explained in the topology file.
Due to a small design flaw (made in December 87), it is today still not
possible to add amino acid 1 to the soup (the only exception is the INIBLD
option, see below). This means that if you want
to build an amino acid N-terminal from the first residue, or if you want
to start building a new protein, you are in trouble. The solution is to
start with reading a very small PDB file (eg with only one GLY in it), or
use the INIBLD option first.
Alternatively, you use SOUCOP in the SOUP menu to duplicate the molecule
in the soup, build C-terminal of the second one, and delete the first
molecule after you finished building N-terminal.
Sorry for this...
The command INIBLD will cause WHAT IF to prompt you for a residue type.
It will then create a new molecule in the soup (in case there are more molecules,
this new molecule will be the last protein, just before the non-protein
molecules) consisting of that one residue. Use the command CBLDS to keep
adding residues to this one if you want to build a
whole molecule. See the parameters about building a certain kind of
secondary structure.
The command NBLD will cause WHAT IF to prompt you for the number of the
residue before which to insert, and the residue type to be inserted. If
the residue before which you want to insert is not the N-terminal residue of
the molecule in which it sits
you will get a warning, and a question about continuing or not.
WHAT IF will then add the new residue at the N-terminal side of the given
residue number.
See the parameters about building a certain kind of secondary structure.
The command CBLD will cause WHAT IF to prompt you for the number of the
residue after which to insert, and the residue type to be inserted. If
the residue after which you want to insert is not the C-terminal residue of
the molecule in which it sits
you will get a warning, and a question about continuing or not.
WHAT IF will then add the new residue at the C-terminal side of the given
residue number.
See the parameters about building a certain kind of secondary structure.
The command CBLDS will cause WHAT IF to prompt you for the number of the
residue after which to insert, Thereafter you will continuously be
prompted for new residue types. These will be added one after the other
C-terminal
of the previous one. Hit zero when prompted for a new residue
if you want to stop adding new residues.
See the parameters about building a certain kind of
secondary structure.
The command FILGAP will cause WHAT IF to prompt you for the number of the
residue after which to fill a gap, Thereafter you will continuously be
prompted for new residue types. These will be added one after the other
C-terminal
of the previous one. Hit zero when prompted for a new residue
if you want to stop adding new residues.
WHAT IF will than try to optimize (rather crudely!) the new residues
such that they fill the gap, i.e. the last inserted residue should connect
well to the first residue after the gap.
See the parameters about building a certain kind of
secondary structure.
At present this option is still a bit buggy.
WARNING. Do NEVER believe anybody who claims he/she can do this.
WHAT IF has a very crude option to build a structure from just a sequence
and the predicted secondary structure. You need to provide a file with
one line per amino acid. The format has to be 1X,A3,1X,A1 thus one blank,
the amino acid in three letter code, one blank, and the secondary
structure. The secondary structure should be one of H, S, 1, 2, C.
H, S, C stands Helix, Sheet, and Coil respectively. The 1 and 2 are the
first and the second residue in a beta-turn respectively.
Do NOT trust this option. The structures it create are wrong. However, at
least all the mechanics for all following options or programs are correctly
done.
This option takes lots of CPU time. The amount
of CPU time needed critically depends on the number of turns in the
secondary structure. If there are N turns, the CPU time will be AA*K to
the power N. AA is the number of amino acids, and K a large constant that
depends on your computer.
In case you have knowledge about CYS-CYS bridges in a protein that either was
made with NEWBLD or otherwise is not very well build, you can
try to optimize these with the OPTCYS option. First use SETCYS in the
SOUP menu to tell WHAT IF which pairs to create, and use SHOHST to tell
WHAT IF which residues are in turns, and then run the OPTCYS
option so often as needed or wanted. This is likely not to be a big improvement,
but then, ab-initio building is no-good anyway.
Be aware that this is also a good option to burn some CPU time.
In case you have an electron density map, you can use the DENFIT option to
optimize the position of a residue in the density. After DENFIT WHAT IF
will prompt you for a residue, the maximal Van der Waals radius overlap
(that is how many Angstrom is an atom allowed to penetrate into another),
the number of steps per chi-angle (=side chain torsion angle) and how the
step size in degrees per chi-angle step. WHAT IF will then brute-forcedly
search the best density fit that can be obtained by rotating the side
chain torsion angles as indicated, thereby keeping the Van der Waals
clashes less than indicated. See also the XRAY menu for options like these.
In case you have an electron density map, you can use the DENFTS option to
optimize the position of a residue in the density. After DENFIT WHAT IF
will prompt you for a residue range, the maximal Van der Waals radius overlap
(that is how many Angstrom is an atom allowed to penetrate into another),
the number of steps per chi-angle (=side chain torsion angle) and how the
step size in degrees per chi-angle step. WHAT IF will then brute-forcedly
search the best density fit that can be obtained by rotating the side
chain torsion angles as indicated, thereby keeping the Van der Waals
clashes less than indicated. See also the XRAY menu for options like these.
The command PARAMS will bring you to the menu from which you can change the
parameters relevant for the build menu.
The parameter BLDTYP can be used to toggle between default building (using
phi=-180, psi=180, and omega=180) or using the user defined parameters for
phi, psi and omega. In case you want the latter, do not forget to set
phi, psi and omega with the BLDPHI, BLDPSI and BLDOME parameters
respectively.
In case you want to build with user define backbone torsion angles (see
parameter BLDTYP) the parameter BLDPHI can be used to set the
default phi angle.
In case you want to build with user define backbone torsion angles (see
parameter BLDTYP) the parameter BLDPSI can be used to set the
default psi angle.
In case you want to build with user define backbone torsion angles (see
parameter BLDTYP) the parameter BLDOME can be used to set the
default omega angle.
The parameter HELIX can be used to make WHAT IF build residues with the
torsion angles phi=-57, psi=-57, omega=180. This overrules the BLDTYP
flag.
The parameter SHEET can be used to make WHAT IF build residues with the
torsion angles phi=-150, psi=150, omega=180. This overrules the BLDTYP
flag.
The command SETPOS can be used to place secondary structure elements at
any desired position in space. You will be prompted for the residue range.
The line connecting the endpoints of this secondary structure element
will be calculated. You will be prompted for a point and a direction
vector. The secondary structure element will be moved such that its
N-terminal endpoint will coincide with the point, and the line connecting
its endpoints will coincide with the direction vector you gave.
No attempts are made to determine or optimize the rotation around the
vector. This will be random.
Not all commands are immediately active in the BUILD menu. By typing
MORE, more commands will be activated.