ClusterAlign: A fiducial tracking and tilt series alignment tool for thick sample tomography

Area of Science:

• Electron Microscopy
• Structural Biology
• Computational Imaging

Background:

• Thick specimens in cryo-scanning transmission electron tomography (cryo-SXT) present reconstruction challenges due to variable fiducial marker visibility.
• Conventional alignment workflows often require manual intervention and exclusion of data, reducing reconstruction resolution.
• Accurate alignment is crucial for high-resolution 3D reconstruction in cryo-SXT.

Purpose of the Study:

• To develop an automated approach for robust tilt series alignment in cryo-SXT.
• To overcome limitations of conventional alignment methods for thick specimens with challenging fiducial visibility.
• To improve the accuracy and efficiency of 3D reconstruction in cryo-SXT.

Main Methods:

• Introduced ClusterAlign software for tilt series alignment.
• Identified fiducial particle clusters based on coherent rotational transformation relative to an untilted reference.
• Integrated tools for alignment error evaluation via projection matching and compatibility with standard software (IMOD/etomo, tomo3d, tomoalign).

Main Results:

• ClusterAlign demonstrates robustness against poor signal-to-noise ratio and varying fiducial visibility.
• The software successfully aligns tilt series to completion, even where other methods fail.
• Automated alignment reduces the need for manual intervention and data exclusion.

Conclusions:

• ClusterAlign provides an automated and robust solution for cryo-SXT tilt series alignment.
• The method enhances the quality and reliability of 3D reconstructions from challenging specimens.
• This approach facilitates more accessible and accurate structural analysis using cryo-SXT.