BetAware

BetAware predicts and annotates transmembrane β-barrel (TMBB) proteins to detect TMBBs and predict their topology for analysis of outer-membrane β-barrels in Gram-negative bacteria, mitochondria, and chloroplasts.

Key Features:

• Machine Learning-Based Detection: Employs N-to-1 Extreme Learning Machines for TMBB detection, reporting a Matthews correlation coefficient of 0.82, a probability of correct prediction of 0.92, and sensitivity of 0.73.
• Topology Prediction: Predicts TMBB topology to address challenges posed by complex β-barrel structures and their low representation in the PDB.
• Integration with GRHCRFs: Leverages Grammatical-Restrained Hidden Conditional Random Fields (GRHCRFs), an extension of HCRFs, to incorporate grammar rules into discriminative biosequence labeling models.

Scientific Applications:

• Membrane protein research: Identification and annotation of TMBBs to study membrane protein structure and function.
• Microbiology and pathogenicity: Analysis of bacterial outer-membrane β-barrel proteins to investigate roles in pathogenicity.
• Drug and vaccine development: Support for identifying TMBBs as potential targets for vaccine or therapeutic interventions.

Methodology:

Uses N-to-1 Extreme Learning Machines for classification and GRHCRFs (Grammatical-Restrained Hidden Conditional Random Fields) to incorporate grammatical constraints into discriminative sequence labeling.