---- leap/lib/0README

This 0README is copied from the leap build tree to the user tree
along with the residue database .lib files and the parm*.dat force 
field files, so if you are reading it in the user area, be aware
that some of the files referenced are back in the 'original' tree.

-- LEaP scripts

makelibs.cmd	Builds the residue libraries for the all-atom part 
		of the '91 version of Weiner et al. and the '94 
		Cornell et al. force fields.

water.cmd	called by makelibs.cmd - makes single water molecule
ions94.cmd	called by makelibs.cmd - makes monoatomic ions
ions91.cmd	called by makelibs.cmd - makes monoatomic ions
ff94.cmd	called by makelibs.cmd - converts prep.in files
ff91.cmd	called by makelibs.cmd - converts prep.in files

-- resulting libs

water.lib	water molecule & WATBOX216 (built by utilConvertWaterBox)
ions94.lib	monoatomic ions
ions91.lib	monoatomic ions

all_nucleic94.lib
all_amino94.lib
all_aminont94.lib
all_aminoct94.lib
		Cornell et al. residue libraries, for use with parm94.dat

all_nucleic91.lib
all_amino91.lib
all_aminont91.lib
all_aminoct91.lib
		Weiner et al. '91 version residue libraries, for
		use with parm91X.dat

---

This directory is used for building leap database .lib files
(which are in the Object File Format discussed in the manual). 
The files with .cmd suffix are leap scripts for building the 
libs.  The leaprc in this directory has no initialization - it 
is here to preclude a default leaprc loading stuff that may 
interfere with proper lib builds.

Two main sets of residue libs are built - the '91 version of 
Weiner et al. all atom residues, and the all atom '94 Cornell 
et al. residues. Each must be used with the appropriate
parm.dat forcefield file. 

Note that when residues are built, leap complains about non-
integral charges; these messages are only relevant for 
molecules which are built for simulations and which would 
normally contain a mix of residues providing net integral 
charges.

The '91 libs have a special version of the force field file, 
parm91X.dat, which _must_ be used to ensure correct improper 
torsion potentials (it has no 'wild card' impropers that could 
be misapplied by leap). parm91X.dat also has an additional angle, 
HO-OS-CT, for nucleic residue HB, HE termini so that the user 
will not have to manually adjust the terminal O type to OH.
(These issues do not arise with the '94 residues because 1)
the ff was adapted to leap at the outset, and 2) terminal
nucleic acid residues are explicitly defined.)

No united atom residues are used in leap because of unresolved 
discrepancies in energies when comparing with prep/link/edit/parm 
results. These occur because leap applies pattern matching to 
atom types and topology when assigning improper torsion potentials,  
while prep/link/edit/parm assign impropers via 'hard coding' in 
the prep.in files. Possibly these problems could be resolved by 
developing a parm.dat that provided correct matches to the improper 
torsions, so scripts for building united atom and other untested 
residues are put in the untested/ dir for future development.

---
