PeptiHub

PeptiHub curates cancer-related peptides with precise annotations and provides utilities for peptide exploration, classification, and machine learning-based design.

Key Features:

• Curated Database: PeptiHub houses 874 cancer-related peptides (CRPs) evaluated across 10 organism categories, 26 organs, and 438 cell lines.
• Activity Annotations: Each peptide entry is annotated with specific activities including anticancer, no effect, insignificant effect, carcinogen, and others, plus effectiveness metrics such as rate and IC50 under cancer conditions within individual organs.
• 3D Conformations: Each database entry includes accessible 3D conformations obtained either experimentally or through predictive methodology.
• Search Frameworks: Three search options are provided: basic search, advanced search, and BLAST sequence search.
• Feature Extraction: PeptiHub offers 1141 features including aaindex1 (alpha propensities, beta propensity, composition indices, hydrophobicity, physicochemical properties, and other properties), amino acid composition (Amino acid Composition and Dipeptide Composition), and Grouped Amino Acid Composition (Grouped amino acid composition, Grouped dipeptide composition, and Conjoint triad).
• Feature Selection: Users can select the most pertinent features from the 1141 available features to address specific research questions.
• Machine Learning Support: The annotated properties facilitate construction of robust training and testing datasets for machine learning-based peptide design and peptide classification.

Scientific Applications:

• Cancer-Related Peptide Discovery: Researchers can explore CRPs based on their specific activities and effectiveness within individual organs to identify peptides with desired anticancer properties.
• Peptide Classification: The precise activity annotations and 1141 extractable features enable classification of peptides into functional categories such as anticancer, no effect, insignificant effect, or carcinogen.
• Machine Learning-Based Peptide Design: The structured training and testing datasets, combined with selectable feature sets, support development of predictive models for peptide activity and property prediction.
• Structure-Activity Relationship Studies: Availability of 3D conformations alongside activity annotations enables investigation of relationships between peptide structure and biological function.

Methodology:

PeptiHub peptides were curated from scientific literature and annotated with activities, effectiveness metrics, and 3D conformations obtained experimentally or through predictive methodology. Feature extraction includes aaindex1, amino acid composition, and Grouped Amino Acid Composition categories.