A simple technique to classify diffraction data from dynamic proteins according to individual polymorphs.

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Differences in diffraction data from protein crystals can indicate structural variations and molecular dynamics.
• Analyzing mixed-state data is crucial for understanding protein behavior but requires effective partitioning into single-state clusters.
• Mixed-state data arise from proteins adopting multiple conformations within crystals.

Purpose of the Study:

• To develop and demonstrate a method for clustering mixed-state protein crystal diffraction data.
• To extract information about protein dynamics and conformational states from complex datasets.
• To assign individual data sets to their correct locations in conformational space.

Main Methods:

• A multi-factor clustering approach was employed to classify diffraction data sets.
• Independent observables, specifically unit-cell parameters and intensities, were used for clustering.
• The method was applied to mixed-state diffraction data from chymotrypsinogen (ChTg) crystals.

Main Results:

• The clustering method successfully sorted mixed-state diffraction data into distinct clusters.
• The data from ChTg crystals were observed to populate a trajectory representing the conversion to chymotrypsin.
• The approach effectively assigned data sets to specific conformational states.

Conclusions:

• Multi-factor clustering is an effective strategy for analyzing mixed-state protein crystal diffraction data.
• This method enables the extraction of insights into protein dynamics and conformational landscapes.
• The findings demonstrate the potential for uncovering biological polystates from crystallographic data.