PRR-HyPred

PRR-HyPred predicts and classifies pattern recognition receptors (PRRs) into specific PRR families to support functional characterization and studies of innate immunity.

Key Features:

• Two-layer hybrid framework: A first-layer support vector machine (SVM) discriminates PRR versus non-PRR sequences.
• Family classification: A second-layer random forest classifier assigns predicted PRRs to families including Toll-like receptors, retinoic acid-inducible gene-I-like receptors, nucleotide oligomerization domain-like receptors, and C-type lectin receptors.
• Sequence-encoded optimal features: Predictions use sequence-encoded optimal features extracted from input sequences as machine-learning input.
• Performance metrics: Reported 10-fold cross-validation performance: layer 1 accuracy 83.4% and Matthew's correlation coefficient 0.639; layer 2 accuracy 95% and Matthew's correlation coefficient 0.816.

Scientific Applications:

• Innate immunity research: Enables large-scale identification and cataloging of PRRs relevant to innate immune recognition.
• Functional characterization: Supports assignment of PRR sequences to specific families for downstream functional inference.
• Therapeutic and disease research: Facilitates discovery of PRR-related targets and investigation of mechanisms in immunology and disease.

Methodology:

Sequence-encoded optimal features are used as input to a two-layer machine-learning pipeline comprising a support vector machine for PRR/non-PRR discrimination and a random forest classifier for family assignment, with performance assessed by 10-fold cross-validation.