Abstract

MicroRNAs (miRNAs) represent the largest class of small non-coding RNAs with a key role in post-transcriptional regulation of gene expression. Studies about their well-known role in embryonic and adult cell proliferation and differentiation (Ren et al., 2009) have recently been extended by works aiming at analyzing their role in several types of human cancer (Olive et al., 2010). For this reason, it is important to understand specific biological functions and mechanisms through which they are able to ensure cell homeostasis and to control cell cycle, developmental timing and cancer progression. However, this is not a trivial task because of two main reasons: the complexity of rules governing miRNAs functional targeting, that are still far from being completely elucidated, and the uncertainty of computational predictions. On the other hand, experimental validation of all potential miRNA:mRNA interactions is too much expensive and time consuming if it has to be carried out for any possible predicted interaction. More effective tools are necessary to provide reliable predictions also on the basis of the analysis of potential miRNA targeting in the context of functional interaction gene networks. This task cannot be solved by analyzing single interactions between miRNAs and their target genes. Indeed, in the literature there are several examples of cooperative activities, represented as multiple miRNAs binding the same group of target genes in many relevant biological processes (Pio et al., 2013). Although this aspect emphasizes the possible dependencies among different miRNAs (and/or among their target genes), most of existing works on the prediction of miRNA:mRNA interactions have focused on single miRNA:mRNA pairs (see Shirdel et al., 2011 for a review), often by considering only structural features and ignoring possible functional inter-dependencies. Due to the recognized limits of such approaches, recently, we have proposed a machine-learning based method (Pio et al., 2014) for exploiting the contribution of several prediction algorithms, by automatically combining their contribution on the basis of a model learned by exploiting both validated and predicted interactions. Although this method has been shown to be much more effective with respect to single prediction algorithms and with respect to some baseline combination approaches, the adopted strategy still does not exploit the inter-dependencies among the considered miRNAs and mRNAs (i.e., it considers independently each miRNA:mRNA interaction).

Autore Pugliese

• D'elia Domenica

Tutti gli autori

• G. Pio ; M. Ceci ; D. D'Elia ; D. Malerba

Titolo volume/Rivista

Non Disponibile

Anno di pubblicazione

2015

ISSN

Non Disponibile

ISBN

Non Disponibile

Numero di citazioni Wos

Nessuna citazione

Ultimo Aggiornamento Citazioni

Non Disponibile

Numero di citazioni Scopus

Non Disponibile

Ultimo Aggiornamento Citazioni

Non Disponibile

Settori ERC

Non Disponibile

Codici ASJC

Non Disponibile