TODO.txt

For computing asyn:
1. Action to repeat model, and distinguish linker.
2. MMC sampling.
Optional:
3. Planarity restraint for NH. (Absolute value of a dot product with vector perpendicular to the plane?)
4. Plane shift restraint. (one coordinate strictly higher between polymers.)
5. SPARTA restraints as a guide? 
6. Accurate chem shift prediction may need rotamer reconstruction using SCWRL.
7. ShiftX requires 1.8s instead of ~9s that SPARTA+ needs for asyn polymer. But also needs a different output parser.
BATCH mode: ./shiftx 1A input.pdb output.shiftx

CHECK: consistency of fragments with topology.

Profile reconstruct test.
String -> ByteString

Add test for StericClash that shows something :-)

Scoring 832 models took 1m24s (1m21 user time.)
Among these parsing took 33s.

Potentials used:
CS -> TALOS -> Dihedral (instead of Ramachandran)
BB x BB potential for beta-sheet twist
hbond_*

ROSETTA potentials useful for beta-sheet assembly:

hbond_lr_bb                                backbone-backbone hbonds distant in primary sequence
hbond_sr_bb                                backbone-backbone hbonds close in primary sequence
hbond_bb_sc                                sidechain-backbone hydrogen bond energy
hbond_sc                                   sidechain-sidechain hydrogen bond energy

For full atom:
fa_pair                                    statistics based pair term, favors salt bridges

fa_dun                                     internal energy of sidechain rotamers as derived from Dunbrack's statistics

fa_plane                                   pi-pi interaction between aromatic groups, by default = 0