Using a measure of structural variation to define a core for the globins

Area of Science:

• Structural bioinformatics
• Protein family classification
• Molecular evolution

Background:

• Expanding databases of 3D protein structures enable classification into families.
• Statistical analysis of conserved features is possible for large protein families like globins.
• Previous work established probabilistic methods for defining structural cores based on means and variances.

Purpose of the Study:

• To define structural cores within protein families using a probabilistic approach.
• To identify conserved structural elements and analyze their variation.
• To investigate the relationship between structural variation and sequence variation in protein families.

Main Methods:

• Probabilistic representation (means and variances) to define structural cores.
• Ranking atoms by structural variation to identify conserved regions.
• Comparison of structural variability with sequence variability, correcting for databank sampling bias.

Main Results:

• Helices A, B, G, and H in globins form a low-variance structural core.
• These core helices fold early and superimpose with helix-turn-helix repressors.
• Non-core helices are linked to functional differences and encoded separately.
• Structural variation did not correlate with sequence variation.

Conclusions:

• A probabilistic method effectively defines structural cores in protein families.
• Conserved structural cores (helices A, B, G, H in globins) are critical for protein folding.
• Functional divergence in globins is associated with non-core regions.
• Sequence variability is not a reliable predictor of structural variation in protein families.