CTCHFA: Discovery of Circulating Tumor Cells in Metastatic Breast Cancer and Nonmetastatic Cancer by using Novel Hybrid Hierarchical Clustering Algorithm in Firefly Distance

Abstract

Abstract
Blood testing for circulating tumor cells (CTCs) has emerged as one of the highest fields in cancer research. CTC detection are an early gene marker of reaction to systemic therapy, whereas their molecular characterization has a strong field that can be translated to individualized targeted treatments and spare breast cancer (BC) patients from unnecessary and ineffective therapies. Genomic research regarding CTCs monitoring for BC is limited due to the lack of indicative genes for their detection and isolation. Different analytical systems for CTC identification and isolation have been developed in recent years. Among them, developing novel methods for identification of blood resulting from tumor cell invasion and CTC molecular characterization plays a crucial role regarding BC. The aim of the study was to evaluate the significance of CTCs in BC patients. The major aim of this study was to propose a novel hierarchical firefly algorithm (HFA) for classification of CTCs and identification of CTCs especially for blood testing. The proposed clustering method efficiently detected the CTC by separation of tumor samples into hierarchical structure, where the distance similarity between two tumor samples was calculated using FA. Each firefly acts as individual genes from tumor samples and distance was evaluated between the two samples. The highest distance samples was identified as CTC and classified as individual class. An experimentation result was conducted to publicly available BC samples dataset from microarray data samples. This work followed a hybrid HFA clustering method for identification of CTCs in BC. Proposed HFA clustering algorithm make easy and efficient method to the detection of CTCs major genes from BC dataset samples, evaluate the results of proposed HFA clustering method and existing hierarchical clustering method by using classification parameters like Sensitivity (Sen), Specificity (Spe), Precision (Pr), False Positive Rate (FPR), False Negative Rate (FNR) and Classification Accuracy (CA). The results achieved for proposed HFA clustering is 0.9478% for CA, 0.9216% for Pr, and 0.9637%, 0.0363%, 0.0522% in terms of Sen, Spe and FPR, respectively.

Keywords: Biological processes (BP), breast cancer (BC), circulating tumor cells (CTCs), peripheral blood (PB), and hierarchical clustering

Cite this Article
Mythili S, Kumar AVS., Discovery of Circulating Tumor Cells in Metastatic Breast Cancer and Nonmetastatic Cancer by Using Novel Hybrid Hierarchical Clustering Algorithm in Firefly Distance. Journal of Computer Technology & Applications. 2015; 6(1): 9–18p.