Next: References Up: No Title Previous: Discussion

Acknowledgements

We thank Professor L.N. Johnson for providing a stimulating working environment. We are grateful to the Royal Commission for the Exhibition of 1851 and the Royal Society for support. We thank András Fiser for numerous helpful discussions and a critical review of the manuscript, and Drs J. Kuriyan, S.K. Burley, M.J. Adams and M.E.M. Noble for providing coordinates prior to deposition.

Abbreviations

3D, three-dimensional; Ig, immunoglobulin; RMS, root mean square; SH2, src homology 2; SH3, src homology 3; HNF-3, hepatocyte nuclear factor 3; the standard three letter and one letter abbreviations for amino acids are also used throughout.

Figure 1

Molscript (Kraulis, 1991) drawings illustrating example of three types of 3D structural similarities all with Mouse Ig light chain variable domain (2FBJ_Ldomain 1). The type A similarity is between 2FBJ_Ldomain 1 and a Human Ig light chain variable domain; the type B between 2FBJ domain 1 and 2CD4 domain 1; and the type C between 2FBJ_Ldomain 1 and poplar plastocyanin. Equivalent strand regions within the three structures similar to 2FBJ_Lare shown as arrows; non-equivalent regions are shown as trace.


Figure 2

Complete linkage cluster analysis dendrogram derived by a matrix of the sums of the absolute differences between columns in Table 3. The numbers on axis correspond to the minimum sum of absolute differences for each cluster (eg. W and F cluster together with a sum of absolute differences of ).


Figure 3

How structurally derived sequence alignments and protein 3D structures are used to find side-chain to side-chain contacts common to a pair of 3D structures. The sequence alignment is shown in the top of the figure and the two structures in the same (ie. superimposed) orientation are shown in the bottom of the figure. A pair of residues in contact in Protein 1 (,) are equivalent to another pair in Protein 2 ( , ). The interaction R-D in Protein 1 is ``replaced'' in Protein 2 by the interaction F-A.


Figure 4

How the structurally similarity index behaves as a function of the percent sequence identity ().


Figure 5

How the percentage of positions having the same accessibility category behaves as a function of the percent sequence identity ().


Figure 6

How the percentage of positions within structurally equivalent regions behaves as a function of the percent sequence identity ().

Figure 7

How the percentage of positions having the same three state secondary structure assignment behaves as a function of the percent sequence identity ().


Figure 8

How the RMS deviation for equivalent atoms behaves as a function of the percent sequence identity (). Interesting outliers are specified by their PDB four letter code, their chain letter (if any) and a Roman numeral specifying the number of the domain considered (numbered sequentially from the N-terminus).

Figure 9

How the number of residue pairs with side-chains in contact behaves as a function of the sequence length.


Figure 10

How the number of residues with side-chains in an energetically favourable contact behaves as a function of the total number of residue pairs in contact.


Figure 11

How the percentage of residues with side-chains in contact common to both structures behaves as a function of percent sequence identity (). Interesting outliers are specified as described in Figure 8.


Figure 12

How the percentage of residues with - distances Å common to both structures behaves as a function of percent sequence identity ().


Figure 13

How the percentage of residues with side-chains in an energetically favourable contact common to both structures behaves as a function of percent sequence identity (). Interesting outliers are specified as described in Figure 8. The dashed line shows the expected percentage of favourable interactions for convergently related proteins of a similar 3D structure.


Figure 14

How the percentage of complementary changes behaves as a function of . Four pairs with high percentages are shown as described for Figure 8. The dashed line shows the expected percent complementary changes for convergently related proteins of a similar 3D structure.


Figure 15a
Figure 15b
Figure 15c
Figure 15d
Figure 15e

Five examples of complementary changes. Details are described in the text.



Next: References Up: No Title Previous: Discussion


gjb@
Thu Feb 9 18:06:48 GMT 1995