Yes. that is how it should be done. if you have a map then you convert them to Fourier coefficients directly using Fourier transformation.
Only for some particles density maps are positive everywhere. X-ray particles are one of them. For neutron and electron scattering maps could be negative as well as positive. For electron diffraction in a simplified form there is a relationship (and it is used in some programs, including refmac) between X-ray and electron scattering factors. Using the fact that electron density and electrostatic potential are related by Poisson equation Mott-Bethe derived a formula: f_e(s) PropTo (Z-f_X(s))/s^2 Z is nuclear charge and f_x(s) is X-ray form factor (Fourier transformation of spherical form of electron density of an atom. f_x(0) is the total number of electrons. For charged atoms at s=0 Z-f_X(s) is different from 0 and therefore for positively charged atoms it could be very large positive number. and for negatively charged atom it could be very negative number. (There is screening effect related to realignment of charges around charges also, but still negatively charged atoms will have negative scattering factors at low resolution). At higher resolution all scattering factors become positive. So, if you measurement (density) is at low resolution then density map can be negative. It is true for EM as well as electron diffraction. There is an additional complications in EM when you compare maps from reconstruction to the atomic electrostatic potential: maps have been subjected to some filtering procedure. An additional complication in density maps after scattering experiment is that scattered particles can knock out some hydrogens (eventually knocking out some of atoms or group of atoms) and the result could be an accumulation of negative charges and therefore (at low resolution) weakening of maps, in some cases making them negative. There are few more complications (direction dependence, solvent behaviour, dynamic scattering effects etc). Regards Garib P.S. I am sorry that this mail turned out to be longer than I wanted. On 15 Mar 2018, at 14:41, Eleanor Dodson <0000176a9d5ebad7-dmarc-requ...@jiscmail.ac.uk> wrote: > So if I calculate "structure factors" from a map do I just use the map as is > with the negative values included ? > > Eleanor > > On 15 March 2018 at 14:37, Ian Tickle <ianj...@gmail.com> wrote: > > Hi Eleanor > > Electron scattering factors can be negative for negative ions, particularly > at low d*. For a low-resolution map it means that the electric potential is > the opposite sign to what you expected. It's why you often don't see ASP & > GLU side-chains in low resolution EM maps. > > Cheers > > -- Ian > > On 15 March 2018 at 14:17, Eleanor Dodson > <0000176a9d5ebad7-dmarc-requ...@jiscmail.ac.uk> wrote: > I am pig-ignorant about these ,, but this example has negative values as well > as positive.. > > What does this mean? I thought a well phased map would be pretty well all > positive.. > > Eleanor > > Dr Garib N Murshudov MRC-LMB Francis Crick Avenue Cambridge CB2 0QH UK Web http://www.mrc-lmb.cam.ac.uk, http://www2.mrc-lmb.cam.ac.uk/groups/murshudov/