Dear Ed, Eleanor, Pratap and CCP4ers Thank you all for the replies.
In Jane S. Richardson's Alacoil paper (Protein Science, 1995, 4:2252-2260), they defined the vertical offset as follows: choosing the Ca of the "a" heptad position on one helix as the reference zero level and position "a+7" as level 1.0, the relative height of the "a" position on the other helix gives the offset (as a fraction of the seven-residue, two-turn repeat). For parallel coiled coils with helices packing in registry, this offset is 0. I am now working with some anti-parallel tetrameric coiled coils and would like to have a handy program to calculate this. It is okay to calculate the offset in Angstrom as I can then convert it into fractional value. J. Pratap suggested they got an estimation by measuring the mean differences in the distances of the Cbeta positions along the two lines of the hepatdic repeats, which was described in detail in their paper (Calladine et al., 2001 J. Mol. Biol. and Laughton et al 2007 Proteins). Thank you, Pratap. I may also get a reasonable value by following Eleanor's suggestion, as long as I cut the two neighboring helices so that they are equal in length. But still, I am wondering if there is an existing program? Thanks! Jie Edward Berry wrote: > Since Jie Liu is talking about coiled coils and heptad > repeat, I think what may be needed is the displacement > along the bundle axis. So an expression for the best > bundle axis line, and then the projection of different > C-a's onto that line to measure the difference between > them? > > Ed > > Eleanor Dodson wrote: > > > Jie Liu wrote: > > > >>Dear CCP4ers: > >> > >>Does anyone know an existing program to calculate the lateral > >>displacement---the vertical offset, measured as a fraction of the > >>the heptad repeat---of neighboring helices in coiled coils or helix > >>bundles, either parallel or antiparallel? > >> > >>Your input is greatly appreciated. > >> > >>Have a nice weekend! > >> > >>Jie > >> > >> > >> > > > > > > Maybe I dont understand the question.. > > > > If you know what you want to match to what cant you just superpose the > > Cas of helix i onto those of helix j and sort it out from that? > > Selecting the match might be tricky but you must have done that? > > > > You will need to do some geometric fiddling I guess. > > The Superpose molecule option from the GUI would give you the centre of > > mass of each segment and the direction cosines for the transformation, > > which will be those for the helix axes. I think the dot product of the > > COM onto the DCs would give you the relative displacements in As > > > > (X_com1*DC1 + Y_com1*DC2 + Z_com1*DC3) - (X_com2*DC1 + Y_com2*DC2 + > > Z_com2*DC3) > > > > Eleanor
