AlphaFold2 Predicts Alternative Conformation Populations in Green Fluorescent Protein Variants

Area of Science:

• Protein structure prediction
• Computational biology
• Artificial intelligence in biochemistry

Background:

• AlphaFold2 revolutionized protein structure prediction, moving beyond static models.
• Understanding protein dynamics is crucial for function and design.
• Conformational flexibility is key to protein function.

Purpose of the Study:

• To investigate AlphaFold2's capability in characterizing protein dynamics.
• To estimate population ratios of different protein conformations.
• To explore the application of AlphaFold2 in protein design for sensing.

Main Methods:

• Targeted column masking of AlphaFold2's multiple sequence alignment.
• Utilizing AlphaFold2 to predict conformational landscapes.
• Characterizing engineered green fluorescent proteins with alternative β-strands.

Main Results:

• AlphaFold2 successfully characterized two main conformations of engineered green fluorescent proteins.
• The method enabled estimation of the population ratio of these conformations.
• The approach demonstrated AlphaFold2's ability to partially sample conformational landscapes.

Conclusions:

• Targeted masking in AlphaFold2 facilitates the study of protein dynamics and conformational populations.
• This method can accelerate the computational design of proteins for sensing applications.
• AlphaFold2 offers new avenues for understanding and engineering protein function.