iDVIP

iDVIP predicts viral integrase inhibitory peptides (VINIPs) from peptide sequences to identify candidates that inhibit retroviral integrase and block integration of viral DNA into host genomes.

Key Features:

• VINIP prediction: Identifies and predicts viral integrase inhibitory peptides (VINIPs) based on peptide sequence characteristics.
• Integrase targeting: Focuses on peptides that target and block integrase proteins involved in retroviral DNA integration.
• Machine learning classification: Employs an in silico machine learning approach for peptide classification.
• Hybrid feature set: Leverages a hybrid feature set to enhance predictive accuracy and reliability.
• Cross-validation performance: Evaluated by 5-fold cross-validation on a training dataset with sensitivity 85.82%, specificity 88.81%, accuracy 88.37%, balanced accuracy 87.32%, and Matthews correlation coefficient 0.64.
• Independent testing: Demonstrates consistent predictive performance on independent testing.

Scientific Applications:

• Antiretroviral candidate identification: Supports identification of novel peptide candidates that inhibit retroviral integrase.
• Characterization of antiretroviral agents: Facilitates characterization of potential antiretroviral peptides for therapeutic development.
• Retroviral lifecycle studies: Aids studies aiming to impede retroviral replication by blocking integrase-mediated DNA integration.

Methodology:

Uses an in silico machine learning approach with a hybrid feature set, evaluated by 5-fold cross-validation on a training dataset and validated via independent testing.