Fri Aug  6 10:03:24 PDT 2004
T0265

DUE 28 Aug 2004


Fri Aug  6 17:39:09 PDT 2004 Kevin Karplus

Looks like a comparative model with 1lj9A as the template (a.4.5.28).
The t04 search concurs, though the hits are ordered somewhat differently.

We should make a dimer out of this.  When I tried making the dimer
from 1lj9 A+B, the result looks nothing like 1lj9.  The
interdigitating helices of 1lj9 are not present, probably because
T0265 doesn't have the c-terminal helix of 1lj9A.
We need to find a better template for the dimer.  Maybe we should try
running try1 though VAST to get a structural hit?

I don't think we want to use the rr constraints on this target.



Fri Aug  6 19:16:13 PDT 2004 Kevin Karplus

I tried 1bjaA+B as a template for the dimer also, and it looks a
litte better than 1lj9A+B.

Maybe I should try polishing up the monomer a little and then seeing
what can be done to make a decent dimer.

For try2, I'll adjust the weights a bit to encourage better packing
and remove extraneous constraints.  I'll run try2 from the alignments,
but we'll need a polishing run after that that starts from the
existing models.



Sun Aug  8 16:35:41 PDT 2004 Kevin Karplus

The try2 cost function likes try1-opt2 better than try2-opt2.  Rosetta
also likes try1-opt2.repack-nonPC a bit better than
try21-opt2.repack-nonPC.

All the models from alignment are essentially the same, at least for
the first 4 models.  The fifth model has the N-terminal helix swung
180 degrees.

try1 and try2 can be seen to be very similar when superimposed.

Given that the templates are all very similar, our best bet at this
point is probably to ignore the dimerization and just polish the
existing models.


Sun Aug  8 19:20:39 PDT 2004 Kevin Karplus

try3-opt2 looks pretty good, and scores best with the try3.costfcn,
which is unconstrained.

try3-opt2.repack-nonPC also scores best with the Rosetta energy function.

I'm submitting
	try3-opt2
	try3-opt2.repack-nonPC
	try1-opt2
	T0265-1lj9A-t2k-local-str2+CB_burial_14_7-0.4+0.4-adpstyle5
	T0265-1jgsA-t2k-local-str2+CB_burial_14_7-0.4+0.4-adpstyle5

Thu Nov 18 23:47:56 PST 2004	Martina Koeva

Based on the smooth gdt scores:

best sam-t04	49.2688 (try-opt1-scrwl)
best submit	48.2386 (model3, also full auto)
model1		47.7145
auto		48.2390 (also model3)
align		47.4876
robetta best	50.2154 (robetta model1)
robetta1	50.2154