il6pred

il6pred predicts Interleukin-6 (IL-6) inducing peptides to identify peptide sequences that induce IL-6 production for studies of inflammatory responses and COVID-19 vaccine and therapeutic research.

Key Features:

• Dataset: Trained on 365 experimentally validated IL-6 inducing peptides and 2991 non-inducing peptides sourced from the immune epitope database.
• Feature extraction: Computed 9149 peptide features using Pfeature.
• Feature selection: Reduced feature space from 9149 to 186 features using the SVC-L1 technique.
• Top features: Selected the top 10 features with highest classification ability for model development.
• Machine learning methods: Evaluated multiple machine learning algorithms and identified a Random Forest model as the best performer.
• Performance: Random Forest achieved AUROC of 0.84 on the training dataset and 0.83 on an independent validation set.

Scientific Applications:

• SARS-CoV-2 peptide identification: Identifies IL-6 inducing peptides within SARS-CoV-2 proteins to support COVID-19 vaccine research.
• Inflammation and therapeutic research: Facilitates studies of IL-6–mediated inflammation and the development of potential therapeutic strategies.

Methodology:

Models were trained on 365 IL-6 inducing and 2991 non-inducing peptides from the immune epitope database; 9149 features were computed with Pfeature, reduced to 186 via SVC-L1, the top 10 features were used for model development, and machine learning models (best: Random Forest) were evaluated yielding AUROC 0.84 (training) and 0.83 (independent validation).