An approach of encoding for prediction of splice sites using SVM

Huang, Jinyan; Li, T.; Chen, K.; Wu, Juanfang

doi:10.1016/j.biochi.2006.03.006

Cited by 56 publications

(40 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…One of the famous classification-based methods is support vector machine (SVM), which is an accurate and high-performance method [11]. Since the performance of the SVM-based methods largely depends on DNA encoding method, there are some works to effectively encode DNA for feature extraction [12], [13], [14], [15]. Another approach for prediction splice sites is statistical analysis, recently a statistical method is presented for the prediction of donor splice sites, which is based on dinucleotide dependencies at all possible positions [16].…”

Section: Introductionmentioning

confidence: 99%

Evaluating the Accuracy of Splice Site Prediction based on Integrating Jensen-Shannon Divergence and a Polynomial Equation of Order 2

Seyfari¹,

Didehvar²,

Banaee³

et al. 2016

IJCA

View full text Add to dashboard Cite

Advances in DNA sequencing technology have caused generation of the vast amount of new sequence data. It is essential to understand the functions, features, and structures of every newly sequenced data. Analyzing sequence data by different methods could provide important information about the sequence data. One of the essential tasks for genome annotation is gene prediction that can help to understand the features and determine functions of the genes. One of the key steps towards correct gene structure prediction is accurate splice site detection. There are vast numbers of splice site prediction methods, however, a few of them can be incorporated in gene prediction modules because of their complexity. In this paper, a novel model is presented to recognize unknown splice sites in a new genome without using any prior knowledge. Our model is defined based on integrating Jensen-Shannon divergence and a polynomial equation of order 2. Finally, the proposed model is evaluated on Yeast's genome to predict splice sites. The experimental results suggest that the proposed method is an effective approach for splice site prediction.

show abstract

Section: Introductionmentioning

confidence: 99%

Evaluating the Accuracy of Splice Site Prediction based on Integrating Jensen-Shannon Divergence and a Polynomial Equation of Order 2

Seyfari¹,

Didehvar²,

Banaee³

et al. 2016

IJCA

View full text Add to dashboard Cite

show abstract

“…Creating an optimized set of features that best represent the dataset has always remained a challenge for splice site prediction. The presence or absence of certain nucleotide sequences close to the splice sites were considered as features for splice site prediction for a long time [8,9,10,11,12,13,14]. Since all such features were not known, there have been constant efforts to improve or refine features as well as include more relevant features by taking into account recent experimental observations.…”

Section: Introductionmentioning

confidence: 99%

SpliceVec: distributed feature representations for splice junction prediction

Dutta

Dubey

Singh

et al. 2017

Preprint

View full text Add to dashboard Cite

Background: Identification of intron boundaries, called splice junctions, is an important part of delineating gene structure and functions. This also provides valuable insights into the role of alternative splicing (AS) in increasing functional diversity of genes. Identification of splice junctions through RNA-seq is by mapping short reads to the reference genome which is prone to errors due to random sequence matches. This encourages identification of splicing junctions through computational methods based on machine learning. Existing models are dependent on feature extraction and selection for capturing splicing signals lying in vicinity of splice junctions. But such manually extracted features are not exhaustive. Methods: We introduce distributed feature representation, SpliceVec, to avoid explicit and biased feature extraction generally adopted for such tasks. SpliceVec is based on two widely used distributed representation models in natural language processing. Learned feature representation in form of SpliceVec is fed to multilayer perceptron for splice junction classification task. Results: An intrinsic evaluation of SpliceVec indicates that it is able to group true and false sites distinctly. Our study on optimal context to be considered for feature extraction indicates inclusion of entire intronic sequence is better than flanking upstream and downstream region around splice junctions. Further, SpliceVec is invariant to canonical and non-canonical splice junction detection. The proposed model is consistent in its performance even with reduced dataset and class-imbalanced dataset. Conclusions: SpliceVec is computationally efficient and can be trained with user defined data as well.

show abstract

“…Shortly after its introduction, its performance has already either matched or outperformed that of traditional machine learning approaches (e.g., NN) for a wide range of applications including splice sites prediction [2]- [7]. Currently, the SVM approach mainly deals with numerical data (with the exception of special kernel functions), so the DNA sequences must be encoded beforehand in some way.…”

Section: Introductionmentioning

confidence: 99%

Learn from the Information Contained in the False Splice Sites as well as in the True Splice Sites using SVM

Xu¹,

Ma²,

Tao³

2007

Proceedings on Intelligent Systems and Knowledge Engineering (ISKE2007)

View full text Add to dashboard Cite

In splice sites prediction, the information contained in false splice sites is often ignored, which has been recognized to be very valuable. In this paper, three novel encoding approaches, MCM with DTF, MCM with UTF and WAM with UTF, are described, all of which consider the information both in true and false splice sites. From the comparison with 5 other encoding methods, we can conclude: (1) SVM can benefit from the information contained in false splice sites as well as in true splice sites. (2) The performance of MCM with DTF and WAM with DTF is comparative, both of which give the better performance nearly in all cases. (3) The performance of binary vector encoding method is surprisingly good, the potential of which need to be further investigated.

show abstract

An approach of encoding for prediction of splice sites using SVM

Cited by 56 publications

References 14 publications

Evaluating the Accuracy of Splice Site Prediction based on Integrating Jensen-Shannon Divergence and a Polynomial Equation of Order 2

Evaluating the Accuracy of Splice Site Prediction based on Integrating Jensen-Shannon Divergence and a Polynomial Equation of Order 2

SpliceVec: distributed feature representations for splice junction prediction

Learn from the Information Contained in the False Splice Sites as well as in the True Splice Sites using SVM

Contact Info

Product

Resources

About