About IRES:

Translation initiation in eukaryotes occurs principally in a 5'cap-dependent manner. Alternatively, for some mRNA, translation initiation takes place internally in cap-independent manner. An element present in 5'UTR of these mRNAs, called Internal Ribosome Entry Site (IRES), is known to facilitate such internal initiation. Presence of IRES was first experimentally demonstrated in viruses of Picornaviridae family. During viral infections and cytoplasmic stresses, global host translation is inhibited through variety of mechanisms. However, some viral and a subset of cellular transcripts encoding stress responsive and regulatory proteins, manage self-translation mediated by IRES. Therefore, IRES plays an important role in viral replication and development of integrated stress response. IRES elements, thus, holds a great therapeutic importance.

About IRESPred:

IRESPred has been designed considering the following:

  1. Generally housekeeping gene transcripts follow cap dependent initiation pathway whereas, IRES mediated initiation is confined to some viral, and cellular transcripts, which encode regulatory and stress responsive proteins.
  2. Such internal initiation of translation is a consequence of interplay between small subunit ribosomal proteins (SSRPs) and IRES elements.

Therefore, IRESPred uses interaction probabilities of different SSRPs with input 5'UTR sequences along with their features like length, number of upstream AUGs, number of external, internal, hairpin, multi loops, total loops and free energy of predicted secondary structure to perform the prediction of IRES using Support Vector Machine (SVM) based approach. The SVM model has been trained using these features extracted from data set with following composition.

  • Positives: Experimentally proven viral and cellular IRES-Positive 5'UTRs sequences
  • Negatives: CDS and 5'UTRs of housekeeping genes and viral CDS

The IRESPred webserver is is implemented using Apache HTTP server, HTML5, JavaScript5, CSS3, Bootstrap3, PHP and CGI. It is developed by Pandurang Kolekar.