Filling the missing cone in protein electron crystallography

The hyper-resolution property of the Sayre equation is explored for extrapolating amplitudes and phases into the missing cone of data left after tilting a representative protein (rubredoxin) to restricted limits in the electron microscope. At 0.6 nm resolution, a reasonable prediction of crystallographic phases can be made to reconstruct the lost information. Best results are obtained if the goniometer tilt value is greater than approximately +/-60 degrees, but some missing information can be restored if the tilt is restricted to +/-45 degrees.