A New Hidden Markov Model for Protein Quality Assessment Using Compatibility Between Protein Sequence and Structure

Area of Science:

• Computational Biology
• Structural Bioinformatics
• Biophysics

Background:

• Protein structure Quality Assessment (QA) is crucial for predicting and analyzing protein structures.
• The relationship between a protein's amino acid sequence and its 3D structure is fundamental to QA.
• Existing QA methods often struggle to capture the intricate sequence-structure relationship.

Purpose of the Study:

• To develop a novel Hidden Markov Model (HMM) for assessing protein sequence-structure compatibility.
• To enhance the accuracy and reliability of protein structure Quality Assessment (QA).
• To provide both global and local quality scores for comprehensive structure evaluation and refinement guidance.

Main Methods:

• Developed a Hidden Markov Model (HMM) incorporating local structures (angular space), secondary structures, and sequence profiles.
• The HMM encodes local structural information by jointly considering sequence and structure data.
• The model generates global scores for overall structure quality and local scores for specific regions.

Main Results:

• The developed HMM.Z model demonstrates superior performance in protein structure selection compared to state-of-the-art methods (OPUSCA, DFIRE, GOAP, RW).
• HMM.Z achieved better overall selection performance on benchmark datasets.
• The model effectively captures the complex relationship between protein sequence and structure.

Conclusions:

• The novel HMM approach provides an effective method for protein structure Quality Assessment (QA).
• HMM.Z offers improved accuracy in selecting correct protein models.
• The model's ability to provide local scores aids in targeted structure refinement.