This test directory is intended for developers who wish to compare
energies derived from Leap setups to those from the old prep/link/edit/parm
(aka plep) programs.


When comparing sander energies for Leap and plep, the integers at
the beginning of the file (which describe how information is
structured in the 'prmtop' files) can be ignored, e.g.:

28c28
<  NTHETH =     405 MTHETA =     284 NPHIH =     712 MPHIA  =     368
---
>  NTHETH =     405 MTHETA =     284 NPHIH =     777 MPHIA  =     368
30,31c30,31
<  NBONA  =     209 NTHETA =     284 NPHIA =     368 NUMBND =      53
<  NUMANG =     124 NPTRA  =      63 NATYP =      26 NPHB   =      24
---
>  NBONA  =     209 NTHETA =     284 NPHIA =     368 NUMBND =      45
>  NUMANG =      97 NPTRA  =      28 NATYP =      26 NPHB   =      24

Further, when comparing energies, it is important to remember:

  - Leap is double precision, while plep is single (except on 'pure'
    64-bit machines like Cray C90)

  - since atoms can be ordered differently by each 'program',
    the order of evaluation of terms within sander can be
    different

  - For nucleic acids in the '91 ff, the terminal oxygen atoms
    (adjacent to HB and HE hydrogen residues) should technically
    have their types adjusted to OH in Leap, but this creates
    problems with missing parameters, and wasn't considered
    important enough to worry about. This leads to dihedral diffs:

<					DIHED      =      186.7244
--
>					DIHED      =      188.2012

Except for these, energy diffs should be on the order of:

-- sander_aminoan91.out.dif

<  BOND    =        6.2012  ANGLE   =      846.8378  DIHED      =       30.6292
<  VDWAALS =      -50.6483  EEL     =    -1589.1307  HBOND      =       -7.3877
<  1-4 VDW =      449.4116  1-4 EEL =      878.3101  CONSTRAINT =        0.0000
---
>  BOND    =        6.2011  ANGLE   =      846.8372  DIHED      =       30.6335
>  VDWAALS =      -50.6484  EEL     =    -1589.1307  HBOND      =       -7.3877
>  1-4 VDW =      449.4125  1-4 EEL =      878.3102  CONSTRAINT =        0.0000


