A computational methodology to diagnose sequence-variant dynamic perturbations by comparing atomic protein...

Area of Science:

• Structural biology
• Computational biophysics
• Protein dynamics

Background:

• Assessing protein functional health and drug efficacy requires understanding dynamic perturbations caused by amino acid mutations.
• Existing methods may not fully capture subtle dynamic changes induced by mutations.

Purpose of the Study:

• To develop a method for diagnosing dynamic perturbations in proteins using network models of X-ray structures.
• To predict protein functional health and potential drug failure based on mutation-induced dynamics.

Main Methods:

• Utilizing network models of protein X-ray structures to analyze atomic interactions.
• Quantifying the allocation of atomic interactions across 1D, 2D, 3D, and 4D structural levels for amino acids.
• Validating the allocation measure on AB5 toxin variants and Transthyretin mutants.

Main Results:

• Differences in atomic interaction allocation across structural levels accurately recover experimental dynamic perturbations.
• The allocation measure successfully distinguishes dynamic perturbations in structurally robust variants and pathogenic/non-pathogenic mutants.
• Observed changes in the allocation measure for coronaviruses' main proteases suggest potential drug failure despite structural similarity.

Conclusions:

• The proposed network model approach effectively diagnoses mutation-induced dynamic perturbations in proteins.
• This method holds promise for predicting protein functional health and identifying potential drug failures.
• The findings highlight the importance of analyzing atomic interaction allocation for understanding protein dynamics and drug interactions.