Computational prediction of viral miRNAs.

While cloning and/or massive parallel sequencing of small RNAs represent powerful tools for the discovery of novel miRNAs, computational miRNA prediction represents a valuable alternative which can be performed with comparably little technical effort. This is especially true for viruses, as the number of predicted candidates generally remains low and thus within a range that may be readily confirmed by experimental means. Here, we provide a detailed protocol for the prediction of putative miRNA genes using VMir, an ab initio prediction program which we have recently designed specifically to identify pre-miRNAs in viral genomes.