Structural Prediction

[Edwin Wright]

The following statement appeared in Nature:

"There are now many examples, in both functionally similar and dissimilar
proteins, where the sequence similarity is statistically insignificant (less
than 20%) and yet the same topology is observed." (M. Thornton, T.P. Flores,
D.T. Jones, M.B. Swindells, Nature, 354, 105, (1991) - p. 106)

If this situation is common, then it seems that notwithstanding functional
similarity/dissimilarity, topological similarity is dependent on a meta-
sequential order, not readily apparent at the level of independently
displayed sequences.  Presumably, this meta-sequential order is governed by
one of the following possibilities:

1. There is a distinct set of sequences (the basis of which is unknown)
   responsible for a specific topology.

2. The set of sequences responsible for a specific topology are actually
   manifesting conserved positions (albeit non-conserved residues) of a
   single (unknown) meta-sequence.

Is anyone aware of any research with respect to the second possibility?

Regards,
-------
Edwin Wright
85 Spinnaker Drive, A-602
Halifax, Nova Scotia
CANADA B3N 3E3

Telephone: (902) 477-5037

Email: ewright at fox.nstn.ns.ca
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=