A novel artificial neural network method for biomedical prediction based on matrix pseudo-inversion

Cai, Binghuang; Jiang, Xia

doi:10.1016/j.jbi.2013.12.009

Cited by 29 publications

(12 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Artificial neural network is one of the machine learning classification tools which are most widely used in biomedical applications due to good results obtained (Dreiseitl and Ohno-Machado, 2002;Cai and Jiang, 2014;Chen et al, 2015;Shaikhina et al, 2015). It is a nonlinear non-parametric model which can mimic from very simple to very complex problems.…”

Section: Discussionmentioning

confidence: 99%

Analysis of an Electrocardiographic Multilead System by Means of Artificial Neural Networks - Study of Repolarization During Premature Ventricular Stimulation

Torkar

Arini

2016

Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies

View full text Add to dashboard Cite

The ventricular repolarization dispersion (VRD) has been shown to increase with premature stimulation. Moreover, several differences between left ventricular and right ventricular, such as the anatomic properties and fibrillation threshold have been reported. However, few data exist regarding the influence of the site of stimulation on modulation of VRD measure by electrocardiographic. In the present work, several ECG indices of VRD, as a function of the coupling interval and the site of stimulation, were studied in an isolated heart rabbit preparation (n=18), using ECG multilead (5 rows x 8 columns) system with Artificial Neural Networks. In both ventricles, results have shown significant decreases in early repolarization duration, while in the left ventricle we have found significant increases of transmural dispersion. Also, we have observed that when the premature stimuli were applied to the left ventricle, the ventricular repolarization dispersion changes were detected using only one preferential electrode (row1-column3). When stimuli were elicited at the right ventricle, changes of VRD were detected by three electrodes (row3-column1, row2-column1 and row3-column8). Finally, a different ventricular repolarization dispersion was found as a function of the site of stimulation.

show abstract

Section: Discussionmentioning

confidence: 99%

Analysis of an Electrocardiographic Multilead System by Means of Artificial Neural Networks - Study of Repolarization During Premature Ventricular Stimulation

Torkar

Arini

2016

Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies

View full text Add to dashboard Cite

show abstract

“…where x ki is the k th PCP value of the i th segment, b 0 is a constant, and b 1 , ⋯, b k , ⋯, b 531 are the model parameters for each of the 531 PCP features. The solution is obtained by min ∑ i = 1 N ( y i − ŷ l ) 2 , s. t. ∑ k = 0 531 | b k | ≤ s with s being the bound tuning parameter [ 31 – 33 ]. We see that solving the LASSO is a Quadratic Programming (QP) problem.…”

Section: Methodsmentioning

confidence: 99%

“…Logistic Regression (LR) is another widely used regression analysis method based on the logistic function [ 28 ], which has also been popularly employed for prediction and classification in biomedical problems [ 15 , 20 , 29 , 30 ]. Least Absolute Shrinkage and Selection Operator (LASSO) is a shrinkage linear method for regression widely used in prediction and classification, which is simple and can often describe the relation between the inputs and the output adequately and interpretably [ 31 – 33 ].…”

Section: Introductionmentioning

confidence: 99%

“…Two categories (i.e., Bayesian-based and regression-based) machine learning methods are then applied to analyzing the six segment-PCP data sets for ubiquitination site prediction. Bayesian-based methods include Naïve Bayes (NB) [ 19 , 20 , 24 , 34 ], Feature Selection NB (FSNB) [ 20 , 34 ], Model Averaged NB (MANB) [ 34 , 35 ], and Efficient Bayesian Multivariate Classifier (EBMC) [ 20 , 36 ], while regression-based methods are SVM [ 25 ], LR [ 28 ], and LASSO [ 31 – 33 ]. We compared the performance of different prediction models for ubiquitination sites in terms of the values of the Area Under Receiver Operating Characteristic Curve (AUROC).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Computational methods for ubiquitination site prediction using physicochemical properties of protein sequences

Cai

Jiang

2016

BMC Bioinformatics

Self Cite

View full text Add to dashboard Cite

BackgroundUbiquitination is a very important process in protein post-translational modification, which has been widely investigated by biology scientists and researchers. Different experimental and computational methods have been developed to identify the ubiquitination sites in protein sequences. This paper aims at exploring computational machine learning methods for the prediction of ubiquitination sites using the physicochemical properties (PCPs) of amino acids in the protein sequences.ResultsWe first establish six different ubiquitination data sets, whose records contain both ubiquitination sites and non-ubiquitination sites in variant numbers of protein sequence segments. In particular, to establish such data sets, protein sequence segments are extracted from the original protein sequences used in four published papers on ubiquitination, while 531 PCP features of each extracted protein sequence segment are calculated based on PCP values from AAindex (Amino Acid index database) by averaging PCP values of all amino acids on each segment. Various computational machine-learning methods, including four Bayesian network methods (i.e., Naïve Bayes (NB), Feature Selection NB (FSNB), Model Averaged NB (MANB), and Efficient Bayesian Multivariate Classifier (EBMC)) and three regression methods (i.e., Support Vector Machine (SVM), Logistic Regression (LR), and Least Absolute Shrinkage and Selection Operator (LASSO)), are then applied to the six established segment-PCP data sets. Five-fold cross-validation and the Area Under Receiver Operating Characteristic Curve (AUROC) are employed to evaluate the ubiquitination prediction performance of each method. Results demonstrate that the PCP data of protein sequences contain information that could be mined by machine learning methods for ubiquitination site prediction. The comparative results show that EBMC, SVM and LR perform better than other methods, and EBMC is the only method that can get AUCs greater than or equal to 0.6 for the six established data sets. Results also show EBMC tends to perform better for larger data.ConclusionsMachine learning methods have been employed for the ubiquitination site prediction based on physicochemical properties of amino acids on protein sequences. Results demonstrate the effectiveness of using machine learning methodology to mine information from PCP data concerning protein sequences, as well as the superiority of EBMC, SVM and LR (especially EBMC) for the ubiquitination prediction compared to other methods.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-0959-z) contains supplementary material, which is available to authorized users.

show abstract

“…It means simply by using 2 SSR markers, bnlg1347 and bnlg381, 401 drought tolerance inbred lines can be predictable. 402 Support Vector Machine are widely used in computational biology including 403 genomics, proteomics, metabonomics due to their high accuracy, their ability to deal 404 with high-dimensional datasets, and their flexibility in modeling diverse sources of 405 data (Cai and Jiang 2013, O'Fallon et al 2013, Verma and Melcher 2012, Xie et al 406 2009. SVM established two-class classification-based Machine Learning methods 407 can then be applied for developing an artificial intelligence system to classify a new 408 allele or fragment into the member or non-member class.…”

Section: Clustering 381mentioning

confidence: 99%

Molecular Predicting Drought Tolerance in Maize Inbred Lines by Machine Learning Approaches

Beiki

2019

Preprint

View full text Add to dashboard Cite

Learning Approaches 2Drought is one of the prime abiotic stresses in the world. Now, amongst the new 3 technologies available for speed up the releasing of new drought tolerance 4 genotypes, there is an emanate discipline called machine learning. The study 5 presents Machine Learning for identification, classification and prediction of drought 6 tolerance maize inbred lines based on SSR genetic markers datasets generated from 7 PCR reactions. A total of 356 SSR reproducible fragment alleles were detected 8 across the 71 polymorphic SSR loci. A dataset of 12 inbred lines with these fragments 9 prepared as attributes and was imported into RapidMiner software. After removal of 10 duplicates, useless and correlated features, 311 feature attributes were polymorphic, 11 ranging in size from 1500 to 3500 bp. The most important attribute fragment alleles in 12 different attribute weighting selected. Ten datasets created using attribute selection 13 (weighting) algorithms. Different classification algorithms were applied on datasets. 14 These can be used to identify groups of alleles with similar patterns of expression, 15and are able to create some models that have been applied successfully in the 16 prediction, classification and pattern recognition in drought stress. Some 17 unsupervised models were able to differentiate tolerant inbred lines from susceptible. 18Four unsupervised models were able to produce the different decision trees with root 19 and leaves. The most important attribute alleles almost in all of models were 20 phi033a3, bnlg1347a1 and bnlg172a2 respectively, that can help to identify tolerant 21 maize inbred lines with high precision. 22 23

show abstract

A novel artificial neural network method for biomedical prediction based on matrix pseudo-inversion

Cited by 29 publications

References 27 publications

Analysis of an Electrocardiographic Multilead System by Means of Artificial Neural Networks - Study of Repolarization During Premature Ventricular Stimulation

Analysis of an Electrocardiographic Multilead System by Means of Artificial Neural Networks - Study of Repolarization During Premature Ventricular Stimulation

Computational methods for ubiquitination site prediction using physicochemical properties of protein sequences

Molecular Predicting Drought Tolerance in Maize Inbred Lines by Machine Learning Approaches

Contact Info

Product

Resources

About