A glance into the evolution of template-free protein structure prediction methodologies

Area of Science:

• Computational biology
• Structural bioinformatics
• Protein structure prediction

Background:

• Protein structure prediction computational approaches have been researched for over 20 years.
• Template-based modeling shows success, but template-free modeling, especially for proteins larger than 150 amino acids, still faces challenges.
• Ab initio protein structure prediction methodologies have seen improvements over time.

Purpose of the Study:

• To review major strategies for template-free protein structure modeling.
• To discuss tools developed for each template-free modeling strategy.
• To comment on the progress of ab initio protein modeling, referencing the evolution of the CASP platform.

Main Methods:

• Review of existing literature on computational protein structure prediction strategies.
• Analysis of template-free modeling approaches.
• Discussion of ab initio modeling advancements, including deep learning applications.
• Examination of the historical progression through CASP (Critical Assessment of protein Structure Prediction) data.

Main Results:

• Template-based modeling is highly successful, while template-free methods lag for larger proteins.
• Deep learning approaches are emerging as powerful tools for constructing protein backbone structures from amino acid sequences.
• Significant progress has been made in ab initio protein structure prediction methodologies over time.

Conclusions:

• Template-free modeling remains a challenging area, particularly for large proteins, but advancements are being made.
• Deep learning shows significant potential to enhance ab initio protein structure prediction accuracy.
• The evolution of CASP reflects the ongoing progress and challenges in the field of protein structure prediction.