BacEffluxPred

BacEffluxPred predicts and classifies bacterial efflux proteins involved in antibiotic resistance using a support vector machine-based two-tier system to identify AR-associated efflux proteins and assign them to sub-families (ABC, MFS, RND, MATE).

Key Features:

• Two-Tier Prediction System: The first tier distinguishes AR efflux proteins from non-AR efflux proteins with 85.81% accuracy, and the second tier classifies AR efflux proteins into ABC, MFS, RND, and MATE families with accuracies of 92.13%, 85.39%, 91.01%, and 99.44%, respectively.
• Support Vector Machine (SVM) Algorithms: SVMs are employed as the core classification algorithm for both tiers of prediction.
• Validation and Benchmarking: Performance was evaluated using leave-one-out cross-validation (jackknife) and benchmarking against an independent dataset to assess predictive accuracy.

Scientific Applications:

• Efflux-mediated antibiotic resistance identification: Prediction and family-level classification of efflux proteins to identify proteins contributing to antibiotic resistance.
• Mechanistic studies and therapeutic/diagnostic development: Family assignments and predictions to inform studies of efflux mechanisms and support development of efflux-targeted therapeutics and diagnostics.

Methodology:

Support Vector Machine-based two-tier classification with performance assessed by leave-one-out cross-validation (jackknife) and benchmarking on an independent dataset.