prPred-DRLF

prPred-DRLF predicts plant resistance (R) proteins by encoding amino acid sequences with deep representation learning features and classifying them for phytopathology research.

Key Features:

• Deep Representation Learning: Transforms amino acid sequences into numerical vectors using deep learning-derived embeddings.
• BiLSTM embedding: Applies bidirectional long short-term memory (BiLSTM) embedding to capture contextual sequence information.
• UniRep embedding: Uses unified representation (UniRep) embedding to capture protein-level sequence features.
• Feature Fusion: Fuses BiLSTM and UniRep embeddings to integrate complementary sequence representations.
• Light Gradient Boosting Machine (LGBM) classifier: Inputs fused embeddings into an LGBM classifier to predict plant R proteins.

Scientific Applications:

• R protein identification: Detection and classification of plant resistance (R) proteins from amino acid sequences.
• Phytopathology research: Characterization of components involved in pathogen recognition and plant defense mechanisms.
• Crop improvement: Support for identifying candidate R proteins relevant to the development of disease-resistant crops.

Methodology:

Amino acid sequences are encoded using BiLSTM and UniRep deep representation learning embeddings; the BiLSTM and UniRep embeddings are fused and the fused features are used by a Light Gradient Boosting Machine (LGBM) classifier for prediction.