MPACK step by step
MPACK is simple if proper inputs are created adhering to rules. The user
simply interacts with questions. An interactive session with typical questions
from MPACK is described below]
Start the mpack script by calling : mpack
Type core name of the output file : This will be
used for creating different output files. Use simple names like model1 or
mod1. Do not use extensions or other fancy characters!
Do you have a toplogy file: (y/n) ? Say yes if you would like to
use a topology file to derive approximate constraints. You must adhere to
the topology file format.
Do you have a template to extract constraints from ?
If you have a template to extract constraints from enter ‘y’. Please note the first template
used should correspond to your master alignment.
Name of the seq. align. file ? Enter
your sequence alignment file name. The alignment file must have an extension
*.aln and should be in clustalw
format as explained in the alignment rules.
Enter the name of the pdb file ? Enter the pdb file name,
must have an extension *.pdb and should have
exactly the residues contained in the alignment file.
Enter interval (deg) for PSI/PHI (default 10): Upper/lower
intervals for dihedral angles. You can change the default values.
Enter interval (deg) for OMEGA (default 5):
Upper/lower intervals for OMEGA dihedral angle. Note currently
if the residues don’t match in the alignment and if the template consists
of CISPEP then manually one should edit the OMEGA angle (*ex.aco) appropriately and rerun *diarun.sh script.
Number of distance constraints per atom (default 30 15 12.5) : Specify
the number of constraints you would like to extract per atom. You can specify
higher number if the length of the protein (alignment) is small. Higher number of the constraints will compromise the
execution time of the program so do not exceed more that 30 unless you have
5-15 residue template fragment from which you want to extract constraints.
Enter threshold between upper and lower limits (default 0.5) :
the threshold of upper and lower limits, increase threshold if you like to
have more flexibility (generally the case if the % id between the template
and target is low) and reduce it if the proteins
are highly similar. Most cases the default value is fine.
Do you like to specify fragments ? If
you like to manually specify fragments to be used in the modeling then enter
yes. But it is advised to use the automatic option for specifying the fragments
so type ‘n’.
Do you have another template to extract constraints from (y/n) ?
you have multiple template specify yes. The
start residue number corresponding the master alignment should be noted down.
Enter the starting residue number of the target sequence in the alignment
file ? Enter the starting residue
number. Next sequence of questions will be similar
to the above. If you answered no (by specifying ‘n’) for the question do
you have another template to extract constraints from….. then
Do you have disulphide bonds in the target ? If
you have disulphide bonds (you must have already changed in the alignment
‘C’ to ‘c’, otherwise
you must edit the alignment file and restart the program.
You will also be required to know the residue pairs involved in the
CYSS bonds. If you
answer ‘y’ then
No. of SS bonds ?
Specify number of disulphide bonds
Enter res. In 1 SSBND ? (RES1, RES2) Enter
residues involved in the 1st SSBOND with RES1 <RES2.
Do you want to run DIAMOD ? If
you say "yes" it will prepare input files for
How many structures do you want to generate ? If
you have used an alignment that were highly similar it is sufficient to generate
1 structure. If you have used
approximate constraints or a template with very low similarity, multiple
templates (you can generate just 1 structure, if it fails then generate
more models) specify 20-50 structures. If you
specify more than 1 structure then you may be asked to specify a random
seed number (must be an odd number).
Enter how many cycles of SECODG do you want to run (1 if no SECODG) ?
option will not have any meaning if you generate only one structure. If
you generate multiple structure and have used approximate constraints you
can run multiple cycles of DIAMOD.