Prediction of allosteric communication pathways in proteins

Area of Science:

• Computational biology
• Biophysics
• Protein dynamics

Background:

• Allostery is fundamental to biological processes, involving communication between protein sites.
• Understanding allosteric pathways is key to controlling protein function.

Purpose of the Study:

• To develop a computational model for predicting maximum information transfer paths between active and allosteric protein sites.
• To utilize information theory, specifically mutual information, to quantify information transfer between residues.

Main Methods:

• A Hidden Markov Model approach is used to identify residues facilitating optimal information transfer.
• Mutual information between residue pairs is approximated using a Gaussian model.
• The model's validity is assessed via nonlinear information theory.

Main Results:

• The model successfully predicts allosteric communication pathways in six diverse protein systems.
• Predicted pathways correlate well with experimental data, explaining functional mechanisms.
• Identified pathways suggest potential multi-functionality and pre-existing functional states.

Conclusions:

• The developed model offers a computationally efficient and simple method for predicting allosteric pathways.
• These pathways are vital for comprehending and manipulating protein functionality.
• The findings provide insights into protein communication and regulation.