Mon Jun 26 09:45:57 PDT 2006
T0345
Make started Mon Jun 26 09:47:18 PDT 2006
Running on shaw.cse.ucsc.edu

Mon Jun 26 10:03:15 PDT 2006 Kevin Karplus

Blast gets excellent full-length matches to 2f8aA (68% over 183
residues) and 1gp1A (65% over 183 residues).

This is glutathione peroxidase, which apparently normally has a selenocysteine.

We are told this is a monomer in the unit cell.

Mon Jun 26 22:00:38 PDT 2006 Kevin Karplus

The HMMs like 2f8aA best and 1gp1A next also.
The try1-opt2 model is based mainly on 2f8aA.

Mon Jun 26 22:11:29 PDT 2006 Kevin Karplus

The models from alignment are all in pretty close agreement.

try1-opt2 looks pretty good, but there are some gaps that need to be closed.
	T0345.try1-opt2.pdb.gz breaks before (T0345)Y123 with cost 2.3995
	T0345.try1-opt2.pdb.gz breaks before (T0345)P56 with cost 2.04788
	T0345.try1-opt2.pdb.gz breaks before (T0345)D20 with cost 1.49409
	T0345.try1-opt2.pdb.gz breaks before (T0345)K154 with cost 1.26737
	T0345.try1-opt2.pdb.gz breaks before (T0345)V34 with cost 1.15583

A polishing run might be in order before we give this target further thought.

Mon Jun 26 22:19:34 PDT 2006 Kevin Karplus

try2 started on lopez, with sheet constraints from try1.

Wed Jul 12 22:08:54 PDT 2006 Kevin Karplus

With the unconstrained costfc, try2-opt2 is our best model, beating
our SAM_T06_server_TS1.  The next best server model is ROBETTA_TS3.

Rosetta liked try1-opt2.gromacs.repack-nonPC and
try2-opt2.gromacs.repack-nonPC about equally well.

I should probably do another polishing run from them, as there are
still some bad breaks:
T0345.try1-opt2.gromacs0.repack-nonPC.pdb.gz has 12 breaks
	T0345.try1-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)P56 with cost 1.17807
	T0345.try1-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)V34 with cost 1.12218
	T0345.try1-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)Y123 with cost 1.02975
	T0345.try1-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)D20 with cost 0.669372
T0345.try2-opt2.gromacs0.repack-nonPC.pdb.gz has 8 breaks
	T0345.try2-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)P56 with cost 1.39187
	T0345.try2-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)E73 with cost 1.0214
	T0345.try2-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)Y123 with cost 0.973532
	T0345.try2-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)N33 with cost 0.643668


Wed Jul 12 22:37:56 PDT 2006 Kevin Karplus

try3 started on cheep, based only on the gromacs0.repack-nonPC models.

Wed Jul 12 23:43:30 PDT 2006 Kevin Karplus

Both undertaker and Rosetta like try3 much better, but there are still
some bad breaks:

T0345.try3-opt2.gromacs0.repack-nonPC.pdb.gz has 6 breaks
	T0345.try3-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)V34 with cost 1.03768
	T0345.try3-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)A35 with cost 0.936832
	T0345.try3-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)N33 with cost 0.564819
	T0345.try3-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)D20 with cost 0.558617
	T0345.try3-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)K4 with cost 0.520558
	T0345.try3-opt2.gromacs0.repack-nonPC.pdb.gz breaks before (T0345)R58 with cost 0.51009

T0345.try3-opt2.pdb.gz has 14 breaks
	T0345.try3-opt2.pdb.gz breaks before (T0345)Y123 with cost 0.965082
	T0345.try3-opt2.pdb.gz breaks before (T0345)I12 with cost 0.562173

Wed Jul 12 23:57:07 PDT 2006 Kevin Karplus

I think we've reached the point of diminishing returns here.
I submitted

ReadConformPDB T0345.try3-opt2.gromacs0.repack-nonPC.pdb
ReadConformPDB T0345.try3-opt2.pdb
ReadConformPDB T0345.try2-opt2.pdb
ReadConformPDB T0345.try1-opt2.pdb

InFilePrefix 
ReadConformPDB T0345.undertaker-align.pdb model 1	# 2f8aA