Algorithm of IL5pred

This module explains the details of all the algorithms and methods used to develop models for prediction.

Dataset Used

Main Dataset:

1907 IL-5 inducing peptides and 7759 non-IL-5 inducing peptides

Alternate dataset:

1907 IL-5 inducing peptides and 1907 random peptides
The algorithm of this webserver relies on the following three modules:

Prediction model

"Predict" module provides the facility to distinguish IL-5 inducing peptides from non-inducing peptides. User can provide multiple sequences as input to the server to predict the given peptide is IL-5 inducing/non-inducing. Several machine learning techniques were used to develop prediction models.

Design model

"Design" module is used to create all possible analogs/mutants of the input sequence and identify the best analog which initiates cytokine i.e. IL-5 release. Using various machine learning techniques, it can predict whether mutants can induce IL-5 secretion.

Scan model

We also incorporated three scan modules:
I) Protein Scan: This approach is to identify IL-5 inducing regions in the protein sequence, which can futher removed or altered in the potential subunit vaccine candidates.

II) Motif Scan: This approach is to discover motifs or patterns in IL-5 inducing peptides. In this study, we used a powerful pattern discovery software MERCI.

II) BLAST Scan: “Blast Search” module is based on a alignment-based search method, i.e., Basic Local Alignment Search Tool (BLAST). The input query sequence is searched against the database of known IL-5 inducing peptides. A query sequence is predicted as IL-5 inducer if found match or hit in the database; otherwise, it is predicted as non-IL-5 inducer.