Skip to main content

Table 2 IRES prediction methods

From: Emerging role of tumor-related functional peptides encoded by lncRNA and circRNA

Name Characteristics Website
IRESite [111] It is based on experimental data derived from 68 viruses and 115 eukaryotic cells. It furnishes information on experimental IRES fragments including their nature, function, origin, size, sequence, structure, relative position to the surrounding protein coding region, and so on.
IRESfinder [112] It is a logit model-based forecasting tool based on 19 k-mer parameters. Its accuracy is ~80%. IRESfinder is a standalone script for Python that is applicable to high-throughput screening.
IRESPred [113] It predicts viral and cellular IRES via the Support Vector Machine (SVM). This predictive model integrates 35 features based on the sequence and structural properties of UTRs and their probabilities of interacting with small subunit ribosomal proteins (SSRPs). Its accuracy is ~75.75% accuracy and it had a 0.51 Matthews correlation coefficient (MCC) in blind testing.
VIPS [114] It consists of the RNAL folding, RNA Align, and pknotsRG programs. Evaluations of the UTR, IRES, and virus databases disclosed that it has superior accuracy and flexibility and can predict four different sets of IRES.