Dear Hailiang,
This apparently is not the real physics, since the
electron density has to be positive everywhere (hope I am right).
Yes, you are right when you are talking about the electron density.
You are wrong when you are talking about a Fourier synthesis calculated
always at a finite resolution (it is what you have, is it?), even when the
term F000 is used as suggested.
Such a synthesis MUST have NEGATIVE values due to Fourier series
truncation. Allowing such negative values was an important point at the
beginning of density modification procedures (beginning of 80th) and it was
one of the key moments when developping electron density histograms (see
for example Lunin, 1988, Acta Cryst A). Moreover, these points even contain
some information and can be used for example to identify the macromolecular
region (since the deepest minima are usually close to the highest maxima).
With best regards,
Sacha
