Biomolecular coevolution and its applications: Going from structure prediction toward signaling, epistasis, and...

Area of Science:

• Computational Biology
• Molecular Evolution
• Bioinformatics

Background:

• Biomolecular sequences evolve significantly, yet structures and functions remain conserved.
• Advances in sequencing and 'Biological Big Data' enable detailed evolutionary studies.
• Statistical models can identify residue mutations linked by spatial proximity.

Purpose of the Study:

• To leverage evolutionary data for improved biomolecular structure prediction.
• To apply evolutionary frameworks for inferring biological interactions and landscapes.
• To advance the understanding of molecular evolution and its functional implications.

Main Methods:

• Utilizing sophisticated statistical models to infer residue pair mutations from genomic sequences.
• Integrating predicted spatial adjacencies as constraints in structure prediction workflows.
• Employing evolutionary fitness landscape models to analyze biological interactions.

Main Results:

• Achieved protein and RNA structure prediction accuracies near the experimental resolution limit.
• Demonstrated the utility of evolutionary constraints in structure prediction.
• Successfully inferred signaling interactions, epistasis, and mutational landscapes.

Conclusions:

• Evolutionary analysis, particularly using spatial proximity data, is crucial for accurate biomolecular structure prediction.
• The mathematical framework extends beyond structure prediction to reveal functional insights like signaling and epistasis.
• This integrated approach transforms the study of molecular evolution and its impact on biological systems.