recorded from −45 to 45 or from −30 to 30 degrees to obtain a 7.7 Å resolution
structure [62]. There also seem to be no reason to use higher dose for higher tilts or
record dual tilt tomograms for subtomogram averaging if no preferred orientation of
particles is present.
10.4.5 Image Processing
Image rotation is followed by interpolation of noisy image values, which degrades
the high-resolution details. Tomographic reconstruction involves one or two
interpolations (depending on the used software), subtomogram averaging introduces an additional interpolation step when the particles are rotated for averaging.
Bartesaghi and colleagues introduced a hybrid method going back to the original tilt
series during StA [61]. This elegant approach allows performing only one interpolation when bringing the patch of the micrograph with the particle to the final
average. It also allowed evaluating the phase residuals for each of the tilting angle
and potentially for different imaging conditions. The data recorded on CCD allowed
obtaining an 8.4 Å structure after the processing [61].
10.5 Conclusion and Overview
Electron tomography is a powerful modality that permits structural analysis of
unique objects. For thin material science objects it can reveal atomic resolution
using STEM imaging at a resolution of up to 2.4 Å. The wider use of the available
advanced instrumentation and careful processing with the available algorithms will
reveal new structures at atomic resolution. For cryo-ET the quality of the tomograms will be greatly improved by the common use of phase pates in combination
with the new detectors and energy filters. This will improve the resolution of
tomograms to *2–4 nm depending on the sample thickness, which will be very
useful for cell biological applications. For subtomogram averaging the future will
be in the use of the mentioned instrumentation, however the more important role
will be dedicated to data processing. Correction of the distortions mentioned in this
chapter will be limited by noise, however more structures at near-atomic resolution
will be obtained for some samples.
Acknowledgements I thank Alex Noble, Andrea Nans, Mikhail Eltsov and Daniel Castano-Diez
for valuable discussions and comments on the text. The author is funded by the Sofja Kovalevstaja
Award from the Alexander von Humboldt Foundation and acknowledges support from the Max
Planck Society.
278
M. Kudryashev
structure [62]. There also seem to be no reason to use higher dose for higher tilts or
record dual tilt tomograms for subtomogram averaging if no preferred orientation of
particles is present.
10.4.5 Image Processing
Image rotation is followed by interpolation of noisy image values, which degrades
the high-resolution details. Tomographic reconstruction involves one or two
interpolations (depending on the used software), subtomogram averaging introduces an additional interpolation step when the particles are rotated for averaging.
Bartesaghi and colleagues introduced a hybrid method going back to the original tilt
series during StA [61]. This elegant approach allows performing only one interpolation when bringing the patch of the micrograph with the particle to the final
average. It also allowed evaluating the phase residuals for each of the tilting angle
and potentially for different imaging conditions. The data recorded on CCD allowed
obtaining an 8.4 Å structure after the processing [61].
10.5 Conclusion and Overview
Electron tomography is a powerful modality that permits structural analysis of
unique objects. For thin material science objects it can reveal atomic resolution
using STEM imaging at a resolution of up to 2.4 Å. The wider use of the available
advanced instrumentation and careful processing with the available algorithms will
reveal new structures at atomic resolution. For cryo-ET the quality of the tomograms will be greatly improved by the common use of phase pates in combination
with the new detectors and energy filters. This will improve the resolution of
tomograms to *2–4 nm depending on the sample thickness, which will be very
useful for cell biological applications. For subtomogram averaging the future will
be in the use of the mentioned instrumentation, however the more important role
will be dedicated to data processing. Correction of the distortions mentioned in this
chapter will be limited by noise, however more structures at near-atomic resolution
will be obtained for some samples.
Acknowledgements I thank Alex Noble, Andrea Nans, Mikhail Eltsov and Daniel Castano-Diez
for valuable discussions and comments on the text. The author is funded by the Sofja Kovalevstaja
Award from the Alexander von Humboldt Foundation and acknowledges support from the Max
Planck Society.
278
M. Kudryashev