Multiple subunit fitting into a low-resolution density map of a macromolecular complex using a gaussian mixture...

Area of Science:

• Structural Biology
• Computational Biology
• Biophysics

Background:

• Electron microscopy enables low-resolution 3D density map reconstruction of large biomolecular complexes.
• Atomic resolution models from X-ray crystallography can be fitted into these maps to generate high-resolution complex models.
• Current fitting methods require significant computational resources, necessitating more efficient algorithms.

Purpose of the Study:

• To develop a fast and efficient algorithm for fitting atomic models into low-resolution 3D density maps.
• To enable accurate atomic model reconstruction of large biomolecular complexes.

Main Methods:

• Developed a program, "gmfit," utilizing Gaussian mixture models (GMMs) to approximate shapes of density maps and atomic models.
• Employed analytical integration of GMMs to rapidly calculate fitness, attraction, and repulsion potential energies between subunits.
• Utilized steepest-descent method with random initial configurations and symmetry restraints for optimal subunit placement.

Main Results:

• The gmfit program demonstrated efficient fitting of atomic models into simulated low-resolution density maps (down to 30 Å).
• Accuracy was achieved with sufficient Gaussian distribution functions per subunit and symmetry restraints for complexes with >3 subunits.
• Successfully reconstructed a model of the GroEL/ES complex with low root mean-square deviation (14.7 Å).

Conclusions:

• The gmfit program provides an efficient and accurate method for atomic model fitting into cryo-EM density maps.
• This approach facilitates the high-resolution structural determination of large biomolecular complexes.
• Further improvements in accuracy can be achieved by incorporating additional biological and biochemical experimental restraints.