Nicholas Newell - 2017-04-22

Sorry for posting this here - SourceForge points here as the place to go for help, but it looks an email to jmol-users would have been better. I'll do that, and please feel free to delete this.

I'm using Jmol 14.13, which should use the DSSP v2.0 algorithm to compute secondary structure. The command $show defaultStructureDSSP returns true, so this
should indeed be the case. However, Jmol displays residues 365-368 from 7odc.pdb as loop-type, not ribbon, while DSSP 2.0 computes this stretch as type E, part of a strand that strectches from residue 365 to residue 373. To complicate matters, at residue 369 Jmol does not not start the ribbon at the residue's Nitrogen, as is normal, but at its alpha carbon.

However, if I issue the command $calculate structure and then redisplay, the structure is then shown with a ribbon consistent with DSSP, running from 365 to 373. Can anyone tell me why Jmol doesn't display this stretch according to DSSP in the first place? It's true that a ramachandran angle calculation does show the stretch 365 to 368 as loop-type with no ribbon, so perhaps for some reason Jmol is doing a ramachandran calculation instead of DSSP for these residues, but then still doesn't explain why the ribbon starts at residue 369's alpha carbon instead of its Nitrogen.

It's worth noting that this stretch is somewhat ambiguous in secondary structure. Although it shows strand-like H-bonding with an adjacent strand, the chain bends tightly enough at these residues to fit the definition of a beta turn. Does Jmol override DSSP in such cases?

Any insight is appreciated.

 

Last edit: Nicholas Newell 2017-04-22