This website provides a convenient way to examine data from simulations that were performed recently in our lab. As a means of simplifying conformational space, we represented protein structure using a “five-state” conformational model: each residue (Alanine, in this case) could sample a 30- by 30-degree box with one of five states. (See the figure at right for the conformational states).

Each conformation was tested according to two criteria: hard sphere sterics (do any atoms occupy the same place at the same time) and hydrogen bond satisfaction (can all hydrogen bonds be satisfied by the peptide itself or by a virtual hard-sphere water). Conformations that failed this test, i.e. had a low acceptance ratio, are predicted to be unfavorable in folded protein structures. Furthermore, these conformations are expected to be unfavorable in the unfolded state of proteins as well.

Conformations with an acceptance ratio of 38% (1/e) or less are stored on this server. As you browse the conformations, you can examine occurences of that conformation in real protein structures. Since all of these conformations are predicted to be unfavorable to some extend, they should not occur frequently in protein structures. When they do occur, however, most of the time they can be explained simply by some sort of structural anomaly.

The presence of these conformations are real violations to the Flory Isolated Pair Hypothesis, and some are biologically significant. The fact that HHHE, for example, is sterically unfavorable indicates that folded proteins cannot intermix helix and strand.

The number of unfavorable conformations identified by this simple model is by far too large to publish. This database is designed to satisfy the reader’s curiosity as well as give detailed information for further study of restrictions in conformational space.

Home | Information | Raw Data | Reference | Rose Lab