The refinement models were made by running undertaker with various
cost functions.

The first 3 models had extra constraints to hold the conserved residues
in place according to the corresponding residues in 2oucA:

Constraints:
	R28.NH1		Y84.OH	3.75
	R28.NH1		D26.OD1	2.59
	Y84.OH		D26.OD1	2.59
	R28.NH2		S36.O	2.86
	R28.NH2		E32.O	3.11
	R28.NE		Q86.OE1	5.46
	R28.N		D26.OD2	2.71
	
	Y33.OH		H37.NE2	3.89
	Y33.OH		H37.ND1	4.56
	Y33.OH		N43.N	4.13
	Y33.OH		A41.O	2.85


There were also constraints to retain the sheets and helices in all
the models.

Clash and break costs were gradually increased to make the chain more
ideal without having the protein expand into "protein foam".


Model
1	try10-opt3 < try9-opt3 < chimera-try8-try6
	chimera-try8-try6 is mostly try8-opt3.gromacs0,
	but M1-T4 are copied from try6-opt3
	
2	try9-opt3.gromacs0.repack-nonPC
	after undertaker, gromacs was used to do energy minimization,
	then rosetta was used to repack sidechains (except PRO and CYS).
	[This model has the lowest rosetta energy of all the ones
	rosetta repacked.]	 

3	try8-opt3.gromacs0 < try7-opt3 < init TR389 model

4	try6-opt3 < try5-opt3 < try4-opt3.gromacs0 
		< try3-opt3.gromacs0 < try2-opt3 < init TR389 model

5	try5-opt3.gromacs0.repack-nonPC