README_AD

folding with AutoDock potential:

example usage:

peptide -r 10000x3000 -t 2 apgvgvapgvgv -p Bias=NULL,external=5,con8,2,1.0,external2=4,con8,2,1.0,Opt=1,0.5,0.5,-0.5 -o output.pdb

external=5,con8,2,1.0
This calls the autodock grid maps. It will look for rigidReceptor.*.map. 
con8 is the a file indicating the residues to score, usually just all residue number.
2 is a thermo factor that is an addition to the original temperature.
1.0 calls the reconstruction of side chain during the scoring.

external2=4,con8,2,1.0 
This calls the cyclic procedure to create an artificial peptide bond between first and last residue.
con8 is meaningless right now, but can be used to indicate the location of the cyclic bond in the future.
2 is the thermo factor, and it will overwrite the previous one in external.
1.0 is meaningless right now, but can be used to flag the cyclic bond type, etc.

Opt=1,0.5,0.5,-0.5
This calls the optimizing/docking procedure. Opt=0 is to use regular MC.
The following 3 numbers is the weight of the target energy in the optimizing procedure.
In the example above:
targetE = 0.5 * totalE + 0.5 * externalE + (-0.5) * firstlastE
Note that totalE includes internalE, externalE and firstlastE. So the above weights scale down the internal energy by a factor of 2 and remove the energy between the first and last residues. The default weights are set to be 1, 0, 0.