A New Feature Selection Methodology for K-mers Representation of DNA Sequences

Bosco, Giosuè Lo; Pinello, Luca

doi:10.1007/978-3-319-24462-4_9

Cited by 11 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As such, recently developed alignment-free methods [21] holds a promising approach to study regulatory genome. For feature extraction as spectral representation, several studies have shown significant research on DNA sequences [22], [23] based on the performances of conventional machine learning algorithms.…”

Section: Related Workmentioning

confidence: 99%

Toward Deep Learning Approaches for Learning Structure Motifs and Classifying Biological Sequences From RNA A-to-I Editing Events

Gwak

2019

IEEE Access

View full text Add to dashboard Cite

RNA editing is a post-transcriptional RNA sequence modification that alters the mature RNA sequence from its template DNA sequences. RNA editing events are critical in various biological and biochemical mechanisms, and can expand the transcriptomic and proteomic diversity from altered gene regulation to mutations. A-to-I RNA editing is now being vastly detected and quantified on a global scale and gained much attention. A deeper understanding of this process with insufficient genomic annotations and prior knowledge-based filtering steps, such as high-throughput next-generation sequencing techniques, are needed, in addition to data regarding whether an editing location is located in one of the following three main classes of RNA editing sites: ALU in Alu repetitive elements, REP in non-Alu repetitive elements, and NONREP in non-repetitive regions. This study proposes deep learning approaches to ameliorate these issues by learning motif patterns and identifying regions of A-to-I editing events in biological sequences. Using datasets derived from the public REDI portal, 300,000 editing sites were equally divided and used for learning sequence and structure motifs visualized by convolutional kernels. We explored the RNA editing pattern changes using information of positional and class enrichment of learned motifs in the three aforementioned classes. We demonstrated that these newly investigated approaches using large-scale RNA sequencing data offer excellent classification accuracy with a well-optimized convolutional neural network and recurrent neural network classifiers that obtained average area under curves of 0.960 and 0.962, respectively. The findings further decipher the principles underlying RNA editing events and will facilitate more effective RNA sequencing research.

show abstract

Section: Related Workmentioning

confidence: 99%

Toward Deep Learning Approaches for Learning Structure Motifs and Classifying Biological Sequences From RNA A-to-I Editing Events

Gwak

2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…These sequence alignment methods involve a feature selection stage. Spectral representation of DNA sequences can be used to determine the sequence similarity in order to enhance the classification performance [8,9]. The time computational complexity remains the reason for restricting the use of alignment approaches.…”

Section: Introductionmentioning

confidence: 99%

Hybrid Approach for Taxonomic Classification Based on Deep Learning

Soliman¹,

Abd-Alhalem²,

El‐Shafai³

et al. 2022

Intelligent Automation &Amp; Soft Computing

View full text Add to dashboard Cite

Recently, deep learning has opened a remarkable research direction in the track of bioinformatics, especially for the applications that need classification and regression. With deep learning techniques, DNA sequences can be classified with high accuracy. Firstly, a DNA sequence should be represented, numerically. After that, DNA features are extracted from the numerical representations based on deep learning techniques to improve the classification process. Recently, several architectures have been developed based on deep learning for DNA sequence classification. Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) are the default deep learning architectures used for this task. This paper presents a hybrid module that combines a CNN with an RNN for DNA classification. The CNN is used for feature extraction, and this is followed by a subsampling layer, while the RNN is trained for classifying bacteria into taxonomic levels. Besides, a wavelet-based pooling strategy is adopted in the subsampling layer, because the wavelet transform with down-sampling allows signal compression, while maintaining the most discriminative features of the signal. The proposed hybrid module is compared with a CNN based on Random Projection (RP) and an RNN based on histogram-oriented gradient features. The simulation results show that the hybrid module has the best performance among other ones.

show abstract