Multiclass cancer classification and biomarker discovery using GA-based algorithms

Liu, Jane Jijun; Cutler, Gene; Li, Wuxiong; Zheng, Pan; Peng, Syd S.; Hoey, Tim; Chen, Liangbiao; Ling, Xuefeng B.

doi:10.1093/bioinformatics/bti419

Cited by 164 publications

(111 citation statements)

References 24 publications

Supporting

Mentioning

108

Contrasting

Unclassified

Order By: Relevance

“…These candidate biomarkers were validated further on the completely independent prediction set with a consistent predictive accuracy of 87%. This level of accuracy is better than the accuracy of previously reported multiclass tumor classification biomarkers identified through the analysis of cell lines (54) and challenged against a limited number of tissue samples (11,45,55).…”

Section: Discussionmentioning

confidence: 67%

“…Although this discrimination has enhanced our diagnostic acumen, the identification of biomarkers whose expression is shared by most cancers could serve the general purpose of segregating malignant from benign conditions independently of individual taxonomies (11). The identified universal biomarkers could be added to the pathologist's repertoire for the uncovering of cancer invasion when comprehensive histologic evaluation is not sufficient.…”

Section: Discussionmentioning

confidence: 99%

“…Transcriptional profiling of the NCI-60 cancer cell lines and a limited number of tissue specimens showed that multiclass cancer classification may lead to the identification of biomarkers expressed by different cancer types (11). Thus, the current study was aimed at the identification of common genetic traits associated with aggressiveness, uncontrolled proliferation, and metastatic potential, which could, in turn, be exploited as ubiquitous identifiers of malignancy.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Common Cancer Biomarkers

Zhao²,

et al. 2006

View full text Add to dashboard Cite

There is an increasing interest in complementing conventional histopathologic evaluation with molecular tools that could increase the sensitivity and specificity of cancer staging for diagnostic and prognostic purposes. This study strove to identify cancer-specific markers for the molecular detection of a broad range of cancer types. We used 373 archival samples inclusive of normal tissues of various lineages and benign or malignant tumors (predominantly colon, melanoma, ovarian, and esophageal cancers). All samples were processed identically and cohybridized with an identical reference RNA source to a custom-made cDNA array platform. The database was split into training (n = 201) and comparable prediction (n = 172) sets. Leave-one-out cross-validation and gene pairing analysis identified putative cancer biomarkers overexpressed by malignant lesions independent of tissue of derivation. In particular, seven gene pairs were identified with high predictive power (87%) in segregating malignant from benign lesions. Receiver operator characteristic curves based on the same genes could segregate malignant from benign tissues with 94% accuracy. The relevance of this study rests on the identification of a restricted number of biomarkers ubiquitously expressed by cancers of distinct histology. This has not been done before. These biomarkers could be used broadly to increase the sensitivity and accuracy of cancer staging and early detection of locoregional or systemic recurrence. Their selective expression by cancerous compared with paired normal tissues suggests an association with the oncogenic process resulting in stable expression during disease progression when the presently used differentiation markers are unreliable. (Cancer Res 2006; 66(6): 2953-61)

show abstract

Section: Discussionmentioning

confidence: 67%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Common Cancer Biomarkers

Zhao²,

et al. 2006

View full text Add to dashboard Cite

show abstract

“…In this study, we have combined genetic algorithms (GAs) and all paired Support Vector Machines (SVMs) for multiclass cancer identification are done by Peng et al (2003). The development of microarray-based high-throughput gene profiling has led to the hope that this technology could provide an efficient and accurate means of diagnosing and classifying tumors, as well as predicting prognoses and effective treatments are given by Liu et al (2005). The problem of feature selection is a difficult combinatorial task in machine learning and of high practical relevance, e.g.…”

Section: Literature Surveymentioning

confidence: 99%

Cancer Classification using Hybrid Fast Particle Swarm Optimization with Backpropagation Neural Network

Vimaladevi¹,

Kalaavathi²

2014

International Journal of Advanced Research in Computer and Comm

View full text Add to dashboard Cite

: Cancer any type of malignant growth or tumor, caused by abnormal and uncontrolled cell division: it may spread through the lymphatic system or blood stream to other parts of the body .Cancer classification is known to have the keys for addressing the problems based on cancer diagnosis and drug discovery. The proposed method of DNA microarray technique has made continuous processing thousands of gene expressions possible. Using gene expression data, the researchers have started the performance to explore the possibilities of cancer classification. The various methods have been introduced in recent years with promising results. But there are still a lot of problem which need to be addressed and understood. The performance of combining the genetic algorithm and Hybrid Fast PSO-BPN method is used to solve the optimization problems. Hybrid Fast PSO-BPN method is used to improve the accuracy and better convergence rate of genetic algorithm and this method is better for local search. This proposed method is used to overcome from the problem of computational difficulties occur by ill-condition of the square penalty function. The experimental result shows that this proposed method is better in accurate result with less execution time.

show abstract

“…Taguchi method is used to find the Neural Network structure for feature section [22]. Measures like Information measures, distance measures, dependence measures, accuracy measures, consistency measures are used for evaluating the goodness of features [23][24][25][26][27]. Wrapper methods with Genetic algorithms are used for feature selection [28] [29].…”

Section: Introductionmentioning

confidence: 99%

Feature Selection by Mining Optimized Association Rules based on Apriori Algorithm

Rajeswari¹

2015

IJCA

View full text Add to dashboard Cite

This paper presents a novel feature selection based on association rule mining using reduced dataset. The key idea of the proposed work is to find closely related features using association rule mining method. Apriori algorithm is used to find closely related attributes using support and confidence measures. From closely related attributes a number of association rules are mined. Among these rules, only few related with the desirable class label are needed for classification. We have implemented a novel technique to reduce the number of rules generated using reduced data set thereby improving the performance of Association Rule Mining (ARM) algorithm. Experimental results of proposed algorithm on datasets from standard university of California, Irvine (UCI) demonstrate that our algorithm is able to classify accurately with minimal attribute set when compared with other feature selection algorithms.

show abstract

Multiclass cancer classification and biomarker discovery using GA-based algorithms

Cited by 164 publications

References 24 publications

Common Cancer Biomarkers

Common Cancer Biomarkers

Cancer Classification using Hybrid Fast Particle Swarm Optimization with Backpropagation Neural Network

Feature Selection by Mining Optimized Association Rules based on Apriori Algorithm

Contact Info

Product

Resources

About