Figure 1.
Overview of MUSTANG structural sieving server.
Figure 2.
Screen snapshot of the results of a typical structural alignment, showing the sieving results in a Jmol window, Lesk-Hubbard plot, and structure-based sequence alignment.
Figure 3.
Plot of number of residue correspondences vs. RMSD in each structure.
Eukaryotic proteases (3EST, 1TON, 3RP2, 5CHA) are in blue, prokaryotic (1SGT, 2SGA, 3SGB, 2ALP) in red. The boxes highlight superpositions of sieved structures with their corresponding RMSD. At 0.6 Å, the catalytic triad and its wireframe surface are displayed in dark blue.
Table 1.
Details of Serine proteases used in structural alignment and sieving.
Figure 4.
Performance of ‘sieving’ against manually curated alignments from SISYPHUS [8].
Q scores were computed using the program QSCORE. Colors correspond to the 3 category groups in the SISYPHUS database: Homologous (black), Fold (green) and Fragments (red). We did not observe any clear relationship between Q scores, sieving level and the three classes.
Figure 5.
Lesk-Hubbard plot and aligned structures of d1gt8a4 and d1mo9a1.
A) Plot of number of residue correspondences vs. RMSD in each structure (d1gt8a4 is in green, d1mo9a1 in cyan). Both domains exhibit an identical curve. The boxes highlight superpositions of sieved structures with their corresponding RMSD. Van der Waals representations of residues are, for d1gt8a4 and d1mo9a1 respectively: D481/D353 in red, A198/A54 in yellow, E128/E73 in magenta and T489/M361 in orange. FAD molecules are represented as sticks in dark blue (d1gt8a4) and pale blue (d1mo9a1); B) Structural alignment of the domains with bound FAD as generated by MUSTANG.