StackIL6: a stacking ensemble model for improving the prediction of IL-6 inducing peptides

Area of Science:

• Immunology
• Computational Biology
• Bioinformatics

Background:

• Interleukin-6 (IL-6) release is triggered by pathogen-derived peptides and immune cells, driving inflammation.
• IL-6 inducing peptides serve as potential diagnostic biomarkers for disease severity and as therapeutic targets for immune suppression.

Purpose of the Study:

• To develop an accurate computational model for identifying IL-6 inducing peptides.
• To facilitate the investigation of IL-6 inducing peptide mechanisms and their application in diagnostics and immunotherapy.

Main Methods:

• A novel stacking ensemble model, StackIL6, was developed.
• The model integrates twelve feature descriptors from composition, composition-transition-distribution, and physicochemical properties.
• Five machine learning algorithms (extremely randomized trees, logistic regression, multi-layer perceptron, support vector machine, random forest) were used.

Main Results:

• StackIL6 demonstrated superior performance compared to the existing IL6PRED method.
• The model outperformed its individual baseline components on training and independent test datasets.
• StackIL6 exhibits excellent discrimination and generalization abilities.

Conclusions:

• StackIL6 is a highly effective tool for identifying IL-6 inducing peptides.
• The model's performance supports its utility in diagnostic and immunotherapeutic applications.
• A freely accessible web server for StackIL6 is available to aid research.