How Can I Find The Distances Between Alpha Helices And Beta Strands In Proteins?

Question

[Douglas S. Stones]

I'm attempting to repeat a computational experiment performed in: Ron Milo, et al., Superfamilies of Evolved and Designed Network, Science 303, 1538 (2004). Among other experiments, they look for network motifs in protein structures.

According to their paper, their protein networks were generated from protein data from the Protein Data Bank (e.g. 1A4J). The nodes of the network are the alpha helices and beta strands, and an undirected edge is added between pairs of nodes whenever their distance is less than 10 angstroms.

Question: How can I (easily) determine when two of these components are separated by 10 angstroms (without re-inventing the wheel)? In particular, is there a software package that would allow me readily to deduce this network?

Answer 1

[Chris]

I haven't read the paper, but an obvious way (though not the only one) to calculate the distance between two elements could be to first determine the center mass and then calculate their distances to each other. DSSP is a starting point for secondary structure assignment. While helices are easy to handle, it's tricky to determine which strands belong to which sheet. Although this sounds easy, the programming behind that might involve more effort. But if you want to do that large-scale, there's probably no way around it anyway.

Answer 2

[Michael Kuhn]

I've done something like this in the past with PyMOL. You can use the selection algebra to see which atoms are close enough, and to narrow down the secondary structure elements. However, you somehow need to enumerate the helices/strands. I guess you can also use the selection algebra to identify residues that are next to each other in sequence and have the same secondary structure assignment.