Characterization and prediction of mRNA polyadenylation sites in human genes

Abstract

The accurate identification of potential poly(A) sites has contributed to all many studies with regard to alternative polyadenylation. The aim of this study was the development of a machine-learning methodology that will help to discriminate real polyadenylation signals from randomly occurring signals in genomic sequence. Since previous studies have revealed that RNA secondary structure in certain genes has significant impact, the authors tried to computationally pinpoint common structural patterns around the poly(A) sites and to investigate how RNA secondary structure may influence polyadenylation. This involved an initial study on the impact of RNA structure and it was found using motif search tools that hairpin structures might be important. Thus, it was propose that, in addition to the sequence pattern around poly(A) sites, there exists a widespread structural pattern that is also employed during human mRNA polyadenylation. In this study, the authors present a computational model that uses support vector machines to predict human poly(A) sites. The results show that this predictive model has a comparable performance to the current prediction tool. In addition, it was identified common structural patterns associated with polyadenylation using several motif finding programs and this provides new insight into the role of RNA secondary structure plays in polyadenylation.

Poly(A) motif prediction using spectral latent features from human DNA sequences

“…A number of studies have demonstrated that information from relatively short upstream and downstream sequences of the candidate poly(A) motifs can specify the true poly(A) motifs to a great extent (Ahmed et al , 2009; Akhtar et al , 2010; Chang et al , 2011; Cheng et al , 2006; Graber et al , 1999; Ji et al , 2010; Kalkatawi et al , 2012; Legendre and Gautheret, 2003; Liu et al , 2005; Salamov and Solovyev, 1997; Tabaska and Zhang, 1999). Statistical properties of the surrounding sequences were explored in different species, such as yeast (van Helden et al , 2000), fly (Retelska et al , 2006), Arabidopsis and rice (Ji et al , 2010) and human (Chang et al , 2011; Retelska et al , 2006; Tabaska and Zhang, 1999).…”

“…Statistical properties of the surrounding sequences were explored in different species, such as yeast (van Helden et al , 2000), fly (Retelska et al , 2006), Arabidopsis and rice (Ji et al , 2010) and human (Chang et al , 2011; Retelska et al , 2006; Tabaska and Zhang, 1999). Although significant progress has been made to the accuracy of poly(A) motif predictors, especially in human DNA sequences, such methods are all based on using sophisticated features that require additional efforts to extract and are highly dependent on domain knowledge.…”

Sequence determinants of polyadenylation-mediated regulation

“…Some research efforts employed machine learning and deep learning approaches to prediction of alternative polyadenylation events and classification of sequences as PASs (Cheng et al 2006;Akhtar et al 2010;Chang et al 2011;Gao et al 2018;Leung et al 2018;Bogard et al 2019). The deep learning approaches highlight the usefulness of convolutional neural networks (CNNs) for regulatory genomics and provide valuable predictions for PAS classification and isoform choice.…”

Terminitor: Cleavage Site Prediction Using Deep Learning Models

“…Another angle to address the prediction problem is through machine learning (ML) approaches, and several such models have been proposed. These include more traditional ML methods like support vector machine and Hidden Markov Model, but also the latest deep learning models, such as DeeReCT-PolyA and DeepGSR (23)(24)(25)(26). Deep learning models generally outperform traditional classifiers because they abandon manually selected features.…”