A sequence-profile-based HMM for predicting and discriminating beta barrel membrane proteins

Area of Science:

• Proteomics
• Structural Biology
• Bioinformatics

Background:

• Membrane proteins constitute a significant portion (15-30%) of proteomes, crucial for biological functions.
• Current prediction methods often overlook outer membrane proteins (OMPs) due to their unique structures.
• OMPs, particularly beta barrel membrane proteins, require specialized prediction tools distinct from all-alpha membrane proteins.

Purpose of the Study:

• To develop an accurate computational method for predicting the topology of beta barrel membrane proteins.
• To enhance the identification and annotation of OMPs in proteome-wide analyses.
• To differentiate beta barrel membrane proteins from other protein classes like globular and all-alpha membrane proteins.

Main Methods:

• Developed a novel Hidden Markov Model (HMM) incorporating evolutionary information from multiple sequence alignments.
• The HMM features a cyclic architecture with six distinct state types representing protein topology.
• Employed a dynamic programming algorithm for model optimization and prediction accuracy assessment.

Main Results:

• Achieved a per-residue prediction accuracy of 83% using a jackknife procedure.
• Correctly predicted the topology for seven out of twelve test proteins.
• Demonstrated high selectivity: retaining 84% of known OMPs while rejecting 90% of globular and 90% of all-alpha membrane proteins.

Conclusions:

• The developed HMM model offers a robust and selective approach for predicting beta barrel membrane protein topology.
• This method significantly advances the accurate identification and functional annotation of OMPs.
• The predictor aids in distinguishing OMPs, improving the comprehensiveness of proteomic studies.