Endonuclease PvuII (1PVI) DNA - GATTACAGATTACA
CAP - Catabolite gene Activating Protein (1BER)
DNA - GATTACAGATTACAGATTACA Endonuclease PvuII bound to palindromic DNA recognition site CAGCTG (1PVI) DNA - GATTACAGATTACAGATTACA TBP - TATA box Binding Protein (1C9B)
CAP - Catabolite gene Activating Protein (1BER)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)
TBP - TATA box Binding Protein (1C9B)
 

YASARA / WHAT IFCASP4 Results

The transgenic fold prediction method described here has been applied to eight CASP4 targets, five of which have been submitted. As the development of the algorithms started three weeks before the first CASP target expired, there was no time to collect statistically significant data allowing calibration of the various modeling parameters, the results are thus preliminary. The method is in principle fully automated, human intervention was only used for extending / debugging the program during the CASP period, but not to correct alignments etc.

The pictures on the left side show the optimum superposition (lowest sequence dependent RMSD) between experimental structure (green) and model (red). The quality of the prediction relative to other groups can be judged from the GDT plots on the right side (click for larger images). The model submitted as number 1 is shown in blue, models 2 to 5 in cyan, the other groups in orange. The closer a line is to the bottom and right borders, the better the prediction.

  Click for larger image  
  Fig.1: T99 - 56 residues, 48% sequence identity with closest template, 5.29 A CA RMSD      
         
    Click for larger image  
  Fig.2: T111 - 431 residues, 50% sequence identity with closest template, 2.19 A CA RMSD      
         
    Click for larger image  
  Fig.3: T123 - 160 residues, 61% sequence identity to closest template, 4.3 A CA RMSD      
         
    Click for larger image  
  Fig.4: T125 - 141 residues, 17% sequence identity to closest template, 5.2 A CA RMSD. Three of the models 2 to 5 (cyan lines) were significantly better in this case.      
         
    Click for large image  
  Fig.5: T128 - 203 residues, 55% sequence identity to closest template, 0.928 A CA RMSD