TumorHPD

TumorHPD predicts and facilitates the design of tumor homing peptides (THPs) to identify sequences that target tumors and tumor-associated microenvironments such as tumor vasculature.

Key Features:

• Predictive modeling: Support vector machine (SVM) models predict THP potential using amino acid composition, dipeptide composition, and binary profiles.
• Model performance: The amino acid composition model achieved 86.56% accuracy, the dipeptide composition model achieved 82.03% accuracy, and the binary profile model achieved 84.19% accuracy.
• Residue preference analysis: Analysis of preferred residue types in THPs informed the development of predictive models.

Scientific Applications:

• Peptide therapeutic design: Prioritizes and optimizes candidate peptide sequences for development of tumor-targeting therapeutics.
• Targeting tumor microenvironment: Identifies peptides with propensity to home to tumors and associated structures such as tumor vasculature.
• Experimental guidance: Guides experimental design and peptide synthesis by predicting likely tumor-homing activity.

Methodology:

Analysis of preferred residue types in THPs and development of support vector machine (SVM) models using amino acid composition, dipeptide composition, and binary profile features.