Predicting Human miRNA-like Sequences within Human Papillomavirus Genomes.

BACKGROUND: This study presents a prediction of putative miRNA within several Human Papillomavirus (HPV) types by using bioinformatics tools and a strategy based on sequence and structure alignment. Currently, little is known about HPV miRNAs.

METHODS: Computational methods have been widely applied in the identification of novel miRNAs when analyzing genome sequences. Here, ten whole-genome sequences from HPV-6, -11, -16, -18, -31, -33, -35, -45, -52, and -58 were analyzed. Software based on local contiguous structure-sequence features and support vector machine (SVM), as well as additional bioinformatics tools, were utilized for identification and classification of real and pseudo microRNA precursors.

RESULTS: An initial analysis predicted 200 putative pre-miRNAs for all the ten HPV genome variants. To derive a smaller set of pre-miRNAs candidates, stringent validation criteria was conducted by applying <‒10 ΔG value (Gibbs Free Energy). Thus, only pre-miRNAs with total scores above the cut-off points of 90% were considered as putative pre-miRNAs. As a result of this strategy, 19 pre-miRNAs were selected (hpv-pre-miRNAs). These novel pre-miRNAs were located in different clusters within HPV genomes and some of them were positioned at splice regions. Additionally, the 19 identified pre-miRNAs sequences varied between HPV genotypes. Interestingly, the newly identified miRNAs, 297, 27b, 500, 501-5, and 509-3-5p, were closely implicated in carcinogenesis participating in cellular longevity, cell cycle, metastasis, apoptosis evasion, tissue invasion and cellular growth pathways.

CONCLUSIONS: The novel putative miRNAs candidates could be promising biomarkers of HPV infection and furthermore, could be targeted for potential therapeutic interventions in HPV-induced malignancies.