SCMTHP

SCMTHP identifies and characterizes tumor-homing peptides (THPs) by computing amino-acid propensity scores with a scoring card method to predict tumor-targeting activity.

Key Features:

• Scoring Card Methodology: Uses the Scoring Card Method (SCM) to generate interpretable scores that quantify residue contributions to THP activity.
• Propensity Scores for 20 Amino Acids: Computes propensity scores for all 20 amino acids to capture sequence-level determinants of tumor-homing behavior.
• Benchmark Performance: Reports accuracies of 0.827 and 0.798 on the Main and Small benchmark datasets, respectively.
• Comparative Evaluation: Performance was compared against decision trees, k-nearest neighbor, multi-layer perceptron, naive Bayes, and partial least squares regression.

Scientific Applications:

• Targeted therapy development: Prioritizes peptides with tumor-targeting potential to guide design of targeted cancer therapeutics.
• Peptide-based delivery systems: Supports selection of THPs for use as targeting moieties in drug or nanoparticle delivery.
• Mechanistic inference: Provides residue-level propensity information to explore physicochemical determinants and interactions of THPs.

Methodology:

Implements the Scoring Card Method (SCM) to generate propensity scores for 20 amino acids; evaluated on Main and Small datasets with reported accuracies of 0.827 and 0.798, validated by 10-fold cross-validation and independent tests, and compared to decision trees, k-nearest neighbor, multi-layer perceptron, naive Bayes, and partial least squares regression.