Fri May 23 09:09:03 PDT 2008
T0415
Make started Fri May 23 09:09:22 PDT 2008
Running on cheep.cse.ucsc.edu

Fri May 23 09:09:56 PDT 2008 Kevin Karplus

pdb-blast gets moderate hit to 3bs1A (E-value 0.097)

Fri May 23 12:26:04 PDT 2008 Kevin Karplus

The HMMs agree that 3bs1A is the right template.
There is a small super-secondary structure match (4 strands) for the
second hit (2yzsA), but otherwise there don't seem to be any hits. 

For try2, I'll run the predictions using only the 3bs1A alignments.

Fri May 23 13:34:07 PDT 2008 Kevin Karplus

try2-opt3 looks ok, except for a couple of small breaks.

For try3, I'll polish the existing models, increasing clash and break costs.

Fri May 23 15:12:37 PDT 2008 Kevin Karplus

try3-opt3 looks pretty good, but rosetta prefers repacking the gromacs
optimized version, which undertaker thinks has worse breaks.

I can do another polishing run with soft_clashes turned up, but also
dry6.5, dry8, and dry12, to try to keep things dense.  I'll also take
out a number of the constraints, since density and packing should hold
things together now.

Fri May 23 16:44:30 PDT 2008 Kevin Karplus

try4-opt3 has essentially no clashes or breaks and packs well.
Rosetta still likes T0415.try3-opt3.gromacs0.repack-nonPC better.

The residues Rosetta doesn't like are F18, L53, P77, and H80.
The F18-L53 clash is barely noticed by undertaker (possibly because
some clash threshold is a bit too high, probably F.CD2-L.CD2 or F.CE2-L.CD2).
The P77.CG-H80.CB clash is not noticed at all.  Maybe I should tweak
the clash table a bit.

I'll see if changing to exp2-pdb-cullpdb_pc80_res1.2_R0.2_d070810_chains408-2symm
makes any difference.
These clashes seem to agree more with gromacs, so I'll try an
optimization run using them.

try5-opt3 scores slightly better than try4-opt3, and rosetta now like sbest
try5-opt3.gromacs0.repack-nonPC

Mon May 26 09:50:33 PDT 2008 Kevin Karplus

This target looks essentially done (until we get metaserver results).
Fri May 30 10:47:26 PDT 2008   SAM-T08-MQAO hand QA T0415 Submitted
Fri May 30 10:47:26 PDT 2008   SAM-T08-MQAU hand QA T0415 Submitted
Fri May 30 10:47:26 PDT 2008   SAM-T08-MQAC hand QA T0415 Submitted

Tue Jun  3 11:07:14 PDT 2008 Kevin Karplus
The MQAU costfcn likes SAM-T08-server_TS1 and Zhang-Server_TS2
The MQAC costfcn likes Zhang-Server_TS1 and Zhang-Server_TS3 

Tue Jun  3 11:11:53 PDT 2008 Kevin Karplus

I'm starting metaserver runs on shaw and orcas, with try5.costfcn.

Tue Jun  3 18:22:25 PDT 2008 Kevin Karplus

try5 likes best

	try5-opt3
	MQAU1-opt3	< SAM-T08-server_TS1
	try4-opt3
	MQAC1-opt3	< MULTICOM-REFINE_TS3
	try3-opt3

The differences between models is so small that there isn't much point
to fussing with them any more.  I'll submit these 5.