Tue Jun 29 09:01:26 PDT 2004 T0220 Due 10 Aug Tue Jun 29 12:57:51 PDT 2004 Kevin Karplus T0220 looks like a comparative model, with c.2.1.* as the main hit. I'm not sure yet whether this is a full-length hit or just a domain. Wed Jun 30 00:46:44 PDT 2004 Kevin Karplus Everything is in agreement for residues 109-260, but there are secondary structure disagreements in the earlier part and there is an unpaired beta strand in the later part. This may be a domain break. For try2, I'll try all alignments and just pick up the sheet constraints for domain 109-260. If it doesn't start looking consistent, we may have to break into domains. Unfortunately, I suspect that the "other" domain is a split domain, so searching for it and aligning to it may be difficult. The t2k and t04 alignments seem to have picked out different key residues (though agreeing on some). I'm not sure which alignment is better. The t04 selects template 1nytA more strongly in the str2+CB_burial_14_7 scoring. ------------------------------------------------------------ Subject: T0220 cancelation; acception of late 196,197 predictions From: casp 6 Date: Tue, 06 Jul 2004 18:55:32 -0700 Dear CASP6 participants, 1. Today PDB has released the structure 1omo that has identical sequence (though different annotation) to the CASP6 target - T0220. Therefore we have to cancel another CASP6 target. The crystallographers who submitted T0220 to us are not to be blamed for public release of the structure as it was independently solved by a different experimentalist group. ------------------------------------------------------------ Thu Jul 8 20:57:56 PDT 2004 Kevin Karplus I picked up 1omo.pdb and modified score-all.under to compare our prediction with the 1omoA chain. The try1 models are the best of a bad lot in terms of rmsd---nothing gets better than 15 Ang. We need a better measure of quality than overall rmsd when we are this far off. Wed Sep 22 11:30:43 PDT 2004 Kevin Karplus Using GDT score, our best model is try1-opt1-scwrl 24.1406% Robetta's best is robetta2 22.9688% These are rather bad.