Five hierarchical levels of sequence-structure correlation in proteins.

Area of Science:

• Bioinformatics
• Computational Biology
• Structural Bioinformatics

Background:

• Protein folding is a fundamental biological process.
• Understanding protein folding pathways is crucial for predicting protein structure and function.
• Current bioinformatics approaches offer powerful tools for modeling these complex pathways.

Purpose of the Study:

• To review recent advancements in modeling protein folding pathways using bioinformatics.
• To present statistical models for sequence-structure correlations at various levels of protein structural complexity.
• To discuss the parallels between statistical and theoretical models of protein folding.

Main Methods:

• Development of statistical models for sequence-structure correlations.
• Application of sequence profiles, hidden Markov models (HMMs), and interaction potentials.
• Utilizing specific models like I-sites Library, HMMSTR, HMMSTR-CM, and SCALI-HMM for different structural levels.

Main Results:

• Statistical models are established for short motifs, extended motifs, nonlocal motif pairs, and 3D motif arrangements.
• These models capture sequence-structure relationships at increasing levels of detail.
• The reviewed models provide a framework for analyzing protein folding dynamics.

Conclusions:

• Bioinformatics offers robust methods for modeling protein folding pathways.
• The presented statistical models advance the understanding of sequence-structure correlations.
• These computational tools are valuable for protein structure prediction and analysis.