journal menu
![[Figure 2]](ba0014fig2mag.jpg)
|
|
Figure 2
A example of the use of the SCOP sequence similarity search facility is shown on a Macintosh workstation. The PDB90 database is searched using FASTA (Pearson, 1996  ) with the sequence of the PDB entry 1SRO (S1 RNA-binding domain of polyribonucleotide phosphorylase, PNP), which in 1996 was unpublished and target T0004 in the CASP2 structure prediction experiment (Moult et al., 1997) and is used here to illustrate the utility of the search facility in SCOP in looking for distant relationships. Because the headers of the PDB90 file contain a SCOP classification code (a.b.c.d.e) it is immediately obvious when several sequences from the same superfamily or fold are in the list. In this list (the self hit to 1SRO has been removed) none of the matches have a significant score [the E value must be <0.01 for 99% confidence (Brenner et al., 1998)], however a match to the superfamily 2.26.4 (2 = β class; 26 = OB -fold; 4 = nucleic acid-binding proteins superfamily) is found twice, and is the only one. This is indeed the correct fold for 1SRO and further investigation of this promising lead might well result in many users coming to this conclusion.As well as the page being linked to the SCOP classification, on a correctly configured workstation (see below) clicking on the green icons results in a structure that the sequence match is to being automatically loaded into the molecular viewer program RasMol [written by Roger Sayle (Sayle & Milner-White, 1995)] with the sequence of the unknown mapped onto it according to the alignment. The view shown is for one CSP when the button next to the `Seq-Cons' was clicked. The colouring scheme is: red for identical residues; yellow for similar residues (+ in BLAST alignments); green for dissimilar residues; blue for non aligned parts of the chain. From this it can be seen that the majority of the structure is matched and that there are clusters of conserved residues in the core of the β-barrel. This `instant' homology modelling can be a useful way to discriminate interactive between likely and unlikely matches. Throughout SCOP the green icons are used to display protein structures with classification features highlighted. Information and software to configure a workstation to enable this visualization facility are available for download from the SCOP URL. |
![[Figure 2]](ba0014fig2.jpg)
 | STRUCTURAL BIOLOGY |
ISSN: 2059-7983
