On Position-Specific Scoring Matrix for Protein Function Prediction

Jeong, Jong Cheol; Lin, Xiaotong; Chen, Xue-Wen

doi:10.1109/tcbb.2010.93

Cited by 167 publications

(70 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“… Figure 6 shows the comparison of the accuracy of the proposed method with the previously available classification methods. The maximum average accuracy obtained on dataset 1 using the neural network was 87%, which was 15% more than the previous study [13]. Similarly, the classification accuracy using dataset 2 is presented in Figure 2.…”

Section: Discussionmentioning

confidence: 58%

“…The functional categories included in the experiments from the yeast genome were metabolism, transcription, cellular transport, and transport facilities, while from the human genome the functional categories were globin, trypsin, ras, esterase, lipase, cytochrome, and so forth, [13, 21, 26]. The sequences belonging to these families were downloaded from the UniProt knowledge base (UniProtKB) protein database.…”

Section: Resultsmentioning

confidence: 99%

“…The performance of the proposed technique has been validated on different datasets using performance measure metrics like true positive rate (TPR), false positive rate (FPR), specificity, sensitivity, recall, F-measure, and Mathews correlation coefficient (MCC) [13, 35]. It was observed that the least numbers of sequences were misclassified in our experiments.…”

Section: Resultsmentioning

confidence: 99%

“…The empirical analysis of the results shows that the statistical metric-based feature selection technique extracted the more relevant and informative features from a variable-length protein sequence than the PSSM-based technique [13]. The proposed technique explores dependencies between the features in a subset and classes.…”

Section: Discussionmentioning

confidence: 99%

“…Jeong et al introduced a feature extraction method based on the position specific scoring matrix (PSSM) to extract features from a protein sequence [13]. The PSSM consisted of four components: position, probe, profile, and consensus.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Efficient Feature Selection and Classification of Protein Sequence Data in Bioinformatics

Iqbal

Faye

Belhaouari

et al. 2014

The Scientific World Journal

View full text Add to dashboard Cite

Bioinformatics has been an emerging area of research for the last three decades. The ultimate aims of bioinformatics were to store and manage the biological data, and develop and analyze computational tools to enhance their understanding. The size of data accumulated under various sequencing projects is increasing exponentially, which presents difficulties for the experimental methods. To reduce the gap between newly sequenced protein and proteins with known functions, many computational techniques involving classification and clustering algorithms were proposed in the past. The classification of protein sequences into existing superfamilies is helpful in predicting the structure and function of large amount of newly discovered proteins. The existing classification results are unsatisfactory due to a huge size of features obtained through various feature encoding methods. In this work, a statistical metric-based feature selection technique has been proposed in order to reduce the size of the extracted feature vector. The proposed method of protein classification shows significant improvement in terms of performance measure metrics: accuracy, sensitivity, specificity, recall, F-measure, and so forth.

show abstract

Section: Discussionmentioning

confidence: 58%

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Efficient Feature Selection and Classification of Protein Sequence Data in Bioinformatics

Iqbal

Faye

Belhaouari

et al. 2014

The Scientific World Journal

View full text Add to dashboard Cite

show abstract

APEX2S: A two‐layer machine learning model for discovery of host‐pathogen protein‐protein interactions on cloud‐based multiomics data

Chen

Shen

Wang

et al. 2020

Concurrency and Computation

View full text Add to dashboard Cite

Presented by the avalanche of biological interactions data, computational biology is now facing greater challenges on big data analysis and solicits more studies to mine and integrate cloud-based multiomics data, especially when the data are related to infectious diseases. Meanwhile, machine learning techniques have recently succeeded in different computational biology tasks. In this article, we have calibrated the focus for host-pathogen protein-protein interactions study, aiming to apply the machine learning techniques for learning the interactions data and making predictions. A comprehensive and practical workflow to harness different cloud-based multiomics data is discussed. In particular, a novel two-layer machine learning model, namely APEX2S, is proposed for discovery of the protein-protein interactions data. The results show that our model can better learn and predict from the accumulated host-pathogen protein-protein interactions.

show abstract

Accurate Identification of Human Phosphorylated Proteins by Ensembling Supervised Kernel Self‐organizing Maps

Cui

Ding

2020

Molecular Informatics

View full text Add to dashboard Cite

Protein phosphorylation is a vital physiological process, which plays a critical role in controlling survival differentiation, cell growth, metabolism and apoptosis. The accurate identification of whether a protein will be phosphorylated solely from protein sequence is especially useful for both basic research and drug development. In this study, a new predictor specifically designed for the prediction of human phosphorylated proteins is proposed. The proposed method first train two supervised kernel self‐organizing maps (SKSOMs): one is trained with feature from protein physiochemical composition view, while the other is trained with feature from protein evolutionary information view. Then, the two trained SKSOMs are ensembled to perform the final prediction. Rigorous computational experiments show that the proposed method achieves 78.75 % and 0.561 on ACC and MCC, which are 6.96 % and 12.5 % higher than that of the state‐of‐the‐art predictor. Overall, the study demonstrated a new sensitive avenue to identify human phosphorylated proteins and could be readily extended to recognize phosphorylated proteins for other species.

show abstract

On Position-Specific Scoring Matrix for Protein Function Prediction

Cited by 167 publications

References 41 publications

Efficient Feature Selection and Classification of Protein Sequence Data in Bioinformatics

Efficient Feature Selection and Classification of Protein Sequence Data in Bioinformatics

APEX2S: A two‐layer machine learning model for discovery of host‐pathogen protein‐protein interactions on cloud‐based multiomics data

Accurate Identification of Human Phosphorylated Proteins by Ensembling Supervised Kernel Self‐organizing Maps

Contact Info

Product

Resources

About