Tue Jun 24 09:23:54 PDT 2008 T0468 Make started Tue Jun 24 09:25:06 PDT 2008 Running on moai27.kilokluster.ucsc.edu Tue Jun 24 11:23:01 PDT 2008 Kevin Karplus No good hits. best Evalue 18.3 for 1xyiA Note: some of the same hits are coming up as in T0467. Is this just randomness, or are these targets related? Both are from Bacillus cereus, and so are likely from the same group. I may want to concentrate on templates that are in common between the two, or even make an alignment of the two and use that as a seed for a search. They do seem to have many of the same proteins in the t06 alignments. Tue Jun 24 17:22:51 PDT 2008 Kevin Karplus The T0468.try1-opt3 model looks pretty good from R33 on, but the N-terminal meander seems like it belongs attached to the rest of the sheet. (See T0467/try1-opt3 for comparison) Tue Jul 1 15:57:25 PDT 2008 SAM-T08-MQAO hand QA T0468 Submitted Tue Jul 1 15:57:25 PDT 2008 SAM-T08-MQAU hand QA T0468 Submitted Tue Jul 1 15:57:25 PDT 2008 SAM-T08-MQAC hand QA T0468 Submitted Mon Jul 21 12:02:00 PDT 2008 Kevin Karplus Starting metaserver runs! I'm way behind schedule on this one! Mon Jul 21 13:32:54 PDT 2008 Kevin Karplus MQAC1 based on SAM-T08-server_TS1 MQAU1 based on BAKER-ROBETTA_TS5 Mon Jul 21 13:45:52 PDT 2008 Kevin Karplus The MQAU1 run based on BAKER-ROBETTA_TS5 looks reasonable, but the MQAC1 and try1 runs look like trash. Even the residue-residue contact prediction is fairly consistent with the antiparallel sheet in MQAU1. Naturally, Rosetta likes MQAU1-opt3.gromacs0.repack-nonPC, since it is basically a rosetta model. But the BAKER-ROBETTA_TS5 model does not look like what I was coming up with for the T0467 models, which I had thought would be similar. Let me do two optimization runs with the MQAU1 sheets and helices, one from existing models (try3) and one from alignments (try3). Mon Jul 21 17:39:08 PDT 2008 Kevin Karplus try3 did find something somewhat like MQAU1, though not quite the same. I'm submitting try2-opt3 to VAST to see where the BAKER-ROBETTA model came from. Your VAST Search job was submitted at 07/21/2008 20:40:03(EDT). Request ID: 797710865471598501 Mon Jul 21 17:53:28 PDT 2008 Kevin Karplus Top VAST hits by P-value. PDB C D Ali. Len. SCORE P-VAL RMSD %Id Description 1K1Y A 4 34 6.9 10e-4.4 1.6 5.9 Crystal Structure Of Thermococcus Litoralis 4-Alpha- Glucanotransferase Complexed With Acarbose 1NFI E 2 26 7.8 0.0001 2.4 7.7 I-Kappa-B-AlphaNF-Kappa-B Complex 1NB5 I 40 6.9 0.0002 2.3 12.5 Crystal Structure Of Stefin A In Complex With Cathepsin H 1FY7 A 2 41 6.6 0.0003 2.1 2.4 Crystal Structure Of Yeast Esa1 Histone Acetyltransferase Domain Complexed With Coenzyme A 1A87 A 1 42 7.1 0.0003 1.7 4.8 Colicin N 2GU1 A 1 34 6.5 0.0003 1.4 0.0 Crystal Structure Of A Zinc Containing Peptidase From Vibrio Cholerae 1STF I 38 6.9 0.0003 2.1 10.5 The Refined 2.4 Angstroms X-Ray Crystal Structure Of Recombinant Human Stefin B In Complex With The Cysteine Proteinase Papain: A Novel Type Of Proteinase Inhibitor Interaction 3B95 A 1 29 8.1 0.0003 2.9 6.9 Euhmt1 (Glp) Ankyrin Repeat Domain (Structure 2) 1ECF B 2 48 7.9 0.0005 2.0 2.1 Escherichia Coli Glutamine Phosphoribosylpyrophosphate (Prpp) Amidotransferase 1D0Q A 2 31 6.4 0.0006 1.4 9.7 Structure Of The Zinc-Binding Domain Of Bacillus Stearothermophilus Dna Primase 2AU3 A 2 27 6.3 0.0010 1.2 14.8 Crystal Structure Of The Aquifex Aeolicus Primase (Zinc Binding And Rna Polymerase Domains)ÿ 1DCU A 1 52 7.5 0.0012 1.9 9.6 Redox Signaling In The Chloroplast: Structure Of Oxidized Pea Fructose-1,6-Bisphosphate Phosphatase 1F34 B 1 28 6.2 0.0012 1.2 3.6 Crystal Structure Of Ascaris Pepsin Inhibitor-3 Bound To Porcine Pepsin 3B8C A 4 39 6.2 0.0014 1.8 12.8 Crystal Structure Of A Plasma Membrane Proton Pump by seq ID PDB C D Ali. Len. SCORE P-VAL RMSD %Id Description 1ORV A 2 20 5.0 0.0304 1.5 25.0 Crystal Structure Of Porcine Dipeptidyl Peptidase Iv (Cd26) 2GBC A 2 22 5.0 0.0457 1.9 18.2 Native Dpp-Iv (Cd26) From Rat 2HU9 A 1 17 5.0 0.0321 1.2 17.6 X-Ray Structure Of The Archaeoglobus Fulgidus Copz N- Terminal Domain 2JID A 2 23 5.2 0.0145 1.8 17.4 Human Dipeptidyl Peptidase Iv In Complex With 1-(3,4- Dimethoxy-Phenyl)-3-M-Tolyl-Piperidine-4-Ylamine 1IV7 A 43 6.1 0.0119 2.5 16.3 Crystal Structure Of Single Chain Monellin 1MOL A 44 6.3 0.0064 2.4 15.9 Monellin (Single-Chain, Fused) 1Z03 A 2 40 7.7 0.0421 1.7 15.0 2-Oxoquinoline 8-Monooxygenase Component: Active Site Modulation By Rieske-[2fe-2s] Center OxidationREDUCTION 2AU3 A 2 27 6.3 0.0010 1.2 14.8 Crystal Structure Of The Aquifex Aeolicus Primase (Zinc Binding And Rna Polymerase Domains)ÿ 2QC9 B 27 7.6 0.0154 2.8 14.8 Mouse Notch 1 Ankyrin Repeat Intracellular Domain 2O62 A 2 45 8.0 0.0036 1.8 13.3 Crystal Structure Of Hypothetical Protein (Zp_00105914.2) From Nostoc Punctiforme Pcc 73102 At 1.75 A Resolution 2DK7 A 30 6.0 0.0085 2.3 13.3 Solution Structure Of Ww Domain In Transcription Elongation Regulator 1 2FKI A 30 6.3 0.0340 1.4 13.3 Nmr Structure Of Protein Yjbr From Escherichia Coli; Northeast Structural Genomics Consortium Target Er226 1A8L A 38 6.9 0.0480 1.5 13.2 Protein Disulfide Oxidoreductase From Archaeon Pyrococcus Furiosus 1A8L A 1 38 6.9 0.0480 1.5 13.2 Protein Disulfide Oxidoreductase From Archaeon Pyrococcus Furiosus 1PGU B 5 38 6.8 0.0086 2.3 13.2 Yeast Actin Interacting Protein 1 (Aip1), Se-Met Protein, Monoclinic Crystal Form 2A00 A 39 6.7 0.0393 1.6 12.8 The Solution Structure Of The Amp-Pnp Bound Nucleotide Binding Domain Of Kdpb It looks like there are some pretty good hits there (like 1iv7A, 1molA, 1z03A=1z02A, 2o62A) that should have come up with the HMMs. Did they? Mon Jul 21 20:57:53 PDT 2008 Kevin Karplus I added some of the better VAST matches as MANUAL_TOP_HITS and am running try4 and try5 from alignments with only 17 selected targets (8 from best-scores and 9 from VAST hits). Try4 uses the same costfcn as try3, and will be looking for templates like the BAKER-ROBETTA_TS5 model. Try5 uses a similar costfcn, but with the MQAC1 sheets and helices, so will be looking for models like the SAM-T08-server_TS1 one. Tue Jul 22 12:50:03 PDT 2008 Kevin Karplus The best try4 model (from 1vpmA) does not score well, even with the try4 costfcn. try5-opt3 (from 1iv7A) scores slightly worse than try2-opt3 with the try5 costfcn, but the gromacs optimized model pulls slightly ahead. It may be worth doing a polishing run on try5, to see what it turns into. Tue Jul 22 13:00:51 PDT 2008 Kevin Karplus try6 started to polish try5. Note: try5 does not look anything like MQAC1, despite getting sheet and helix constraints from it. It looks a bit like try2, but the placement of the helix is different and the sheet a little different, particularly in how it curls. try4 seems to mess up some of the strands Wed Jul 23 12:50:06 PDT 2008 Kevin Karplus I have 4 models interesting enough to submit: ReadConformPDB T0468.try2-opt3.gromacs0.repack-nonPC.pdb # < MQAU1-opt3 < BAKER-ROBETTA_TS5 # best rosetta energy ReadConformPDB T0468.try6-opt3.pdb # < try5-opt3.gromacs0 < align(1iv7A?) ReadConformPDB T0468.try3-opt3.pdb # < align(1tupA+1yliA+1bqk+2uurA+1vpmA) ReadConformPDB T0468.try4-opt3.pdb # < align(1vpmA?) Perhaps I should try aligning T0467 and T0468 and using alignments of the favorite model for each to generate a model for the other. Wed Jul 23 13:13:45 PDT 2008 Kevin Karplus I added some targets to the Makefile to make global one-track and three-track alignments of T0467 and T0468, and they agree except on the alignment or not of the initial MET of T0467. I'll do the same in T0467, and see if it comes up with the same alignments. Wed Jul 23 13:36:03 PDT 2008 Kevin Karplus There was some disagreement on the position of the first insertion of 5 residues. I think I may like the T0467-based alignments better. Wed Jul 23 18:15:36 PDT 2008 Kevin Karplus The try7-opt3 model, based on alignment to T0467 models is not very convincing, having a huge gap between I24 and E25, but the Cterminal part (E25 on) does not lok bad, and may be worth combining with other models. Wed Jul 23 20:58:05 PDT 2008 Kevin Karplus I'll try a try2-try7 chimera, with try2-opt3.gromacs0.repack-nonPC for M1-N31, and try7-opt3 for L32-K109 Wed Jul 23 21:10:26 PDT 2008 Kevin Karplus try8 will try to polish the chimera-try2-try7 model, mainly closing the gap between the two pieces and resolving any clashes. I think that try8 will end up being my 5th model. Thu Jul 24 10:56:52 PDT 2008 Kevin Karplus I currently have 5 models that I think may be worth submitting: model 1 is called T0468.try2-opt3.gromacs0.repack-nonPC.pdb model 2 is called T0468.try6-opt3.pdb model 3 is called T0468.try8-opt3.pdb model 4 is called T0468.try3-opt3.pdb model 5 is called T0468.try4-opt3.pdb though the try3 and try4 models are rather junky I don't like the N-terminus of try6, but it's positioning of the C-terminal helix is interesting. It might we worth looking at a try2-try6 chimera, with try2 for M1-N34 and try6 for Y35-K109. Another possibility is M1-F51, R52-K109. Thu Jul 24 11:21:59 PDT 2008 Kevin Karplus try9 will attempt to optimize chimera-try2-N34-Y35-try6 try10 will attempt to optimize chimera-try2-F51-R52-try6 Thu Jul 24 13:26:26 PDT 2008 Kevin Karplus Submitted with comment: The SAM HMMs came up with only weak hits for this taret (best was 1xyiA with E-value 19.3), and those hits did not seem to lead to good full-length models, so I ended up working mainly with metaserver models. The BAKER-ROBETTA_TS5 model turned out to be the one I liked best and used most heavily in creating my models. Since T0467 and T0468 are quite similar sequences, I tried to find similar folds for both targets. I also tried using one of my favorite T0467 models as a template, but I did not get anything very useful that way. Model 1 T0468.try2-opt3.gromacs0.repack-nonPC.pdb # < MQAU1-opt3 < BAKER-ROBETTA_TS5 # best rosetta energy 2 T0468.try9-opt3.pdb # < chimera-try2-N34-Y35-try6 chimera-try2-N34-Y35-try6: M1-N34 T0468.try2-opt3.gromacs0.repack-nonPC.pdb Y35-K109 from T0468.try6-opt3.pdb try6-opt3 < try5-opt3.gromacs0 < align(1iv7A?) 3 T0468.try10-opt3.gromacs0.repack-nonPC.pdb # < chimera-try2-F51-R52-try6 chimera-try2-F51-R52-try6: M1-F51 T0468.try2-opt3.gromacs0.repack-nonPC.pdb R52-K109 from T0468.try6-opt3.pdb 4 T0468.try3-opt3.pdb # < align(1tupA+1yliA+1bqk+2uurA+1vpmA) 5 T0468.try4-opt3.pdb # < align(1vpmA?)