Re: [modeller_usage] Restraints in table 6.1

[Modeller Caretaker <>]

On 3/22/10 2:50 AM, abdullah ahmed wrote:
> I have a few very long (and perhaps naive) questions regarding "Table
> 6.1: List of physical Restraints types"

For reference, this is the table at
http://salilab.org/modeller/9v7/manual/node245.html#21703

>    1. If the Hydrogen bonding restraint is applied, does it
>       automatically incorporate all possible Hydrogen bonds in the file,
>       or do the specific hydrogen bonding atoms have to be specified?
>       (Also could you maybe provide a short piece of code showing how
>       this restraint can be applied?)

This table is not a list of restraints. It is a list of categories into 
which you can place your restraints. Thus, the names are essentially 
meaningless - they are just a convenience for the purpose of weighting 
different groups of restraints.

physical.h_bond is used as the category for restraints between two 
beta-strands generated by secondary_structure.sheet(); see 
http://salilab.org/modeller/9v7/manual/node213.html for an example.

>    2. Are the Phi, Psi, omega, and Chi restraints used to lock the
>       angles of the backbone and the the side chain into specific values
>       or are they used for something else all together?

These are used by Restraints.make() - see 
http://salilab.org/modeller/9v7/manual/node196.html - when building the 
various homology-derived dihedral restraints. See that URL for more 
information.

>    3. What do the accessibility, absposition, dihedral_diff and SAX
>       restraints do?

They are used as the categories for some types of restraints. 
accessibility, absposition and dihedral_diff are typically used as the 
category for the similarly-named features at 
http://salilab.org/modeller/9v7/manual/node97.html. The SAXS category is 
for SAXS restraints, which are experimental and incomplete, and thus not 
currently used.

	Ben Webb, Modeller Caretaker
--
             http://www.salilab.org/modeller/
Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage