iDPPIV-SCM

iDPPIV-SCM predicts DPP-IV inhibitory peptides from amino acid sequence information to identify candidates for dipeptidyl peptidase IV (DPP-IV) inhibition relevant to Type 2 diabetes (T2D) research.

Key Features:

• Scoring Card Methodology: Employs the scoring card method (SCM) integrating amino-acid propensity scores to predict DPP-IV inhibitory activity.
• Superior Performance: Outperformed k-nearest neighbor, logistic regression, decision tree, and support vector machines with improvements of 2–11% in accuracy, 4–22% in Matthew's correlation coefficient (MCC), and 7–10% in area under the curve (AUC).
• Interpretability and Simplicity: Provides an interpretable, scoring-based model that allows straightforward biological interpretation of predictions.
• Molecular Insight: Derives estimated propensity scores for amino acids to reveal molecular determinants that enhance DPP-IV inhibitory potency.

Scientific Applications:

• Diabetes research: Supports identification of peptide candidates targeting DPP-IV for therapeutic strategies in Type 2 diabetes (T2D).
• Peptide screening: Enables sequence-based screening of potential DPP-IV inhibitory peptides prior to synthesis to prioritize experimental validation.

Methodology:

Calculates amino-acid propensity scores from peptide sequences and applies the scoring card method (SCM) to predict the likelihood of DPP-IV inhibition, with model validation against independent datasets.