FT3D: three-dimensional Fourier analysis on small Unix workstations for electron microscopy and tomographic studies

Area of Science:

• Computational biology
• Structural biology
• Image processing

Background:

• Three-dimensional reconstructions are crucial for understanding biological structures.
• Electron microscopy generates projection data requiring complex analysis.
• Existing Fourier analysis tools may not be optimized for modern hardware.

Purpose of the Study:

• Introduce FT3D, a C-language software package for three-dimensional Fourier analysis.
• Detail the capabilities of FT3D for evaluating transforms, correlations, and spectra.
• Highlight the implementation of FT3D to leverage modern workstation resources.

Main Methods:

• Developed FT3D as a self-contained package of tools.
• Utilized the C programming language for Unix workstations.
• Implemented algorithms to efficiently compute direct and inverse 3D Fourier transforms, auto- and cross-correlations, and spectra.

Main Results:

• FT3D successfully evaluates direct and inverse 3D Fourier transforms, correlations, and spectra.
• The library is designed to exploit the capabilities of modern workstations for enhanced performance.
• Performance benchmarks (elapsed times) for various job dimensions and RAM buffers are provided for specific hardware.

Conclusions:

• FT3D provides a robust and efficient solution for three-dimensional Fourier analysis in biological structure reconstruction.
• The software's design optimizes computational performance on contemporary hardware.
• The reported benchmarks offer valuable insights for users planning to employ FT3D in their research.