Detection of Genes in Individual Associated with Laryngeal Cancer using ParalTabs

ABSTRACT

This paper shows the experimental results obtained by applying the classifier parallel scheme of decision tables (ParalTabs) to detection of genes in individuals associated with laryngeal cancer¹

¹ This paper is an extended version of two conference paper that appeared in [1] and [2]. The key additions of this journal version are as follows. First, in Section 3 we use a different way of how to implement the ParalTabs algorithm, using a set of processes instead of threads. Second, in Section 4 we use the same transformation of data used in previous work [2], but the size of data in this paper is different. Finally, this paper contains new experimental results in Section 4, one of this is using ParalTabs for searching genetic Q3 Q2 markers of laryngeal cancer, which is different from those shown in the paper [2].

. The experiments are performed by analyzing a chromosomal database of laryngeal cancer which has the International System for Human Cytogenetic Nomenclature format. ParalTabs is an implementation of decision tables classifiers using the parallel model of multiple instruction and multiple data streams. This classifier uses a set of process that communicates via messages passing and use a parallel scheme that follows the strategy of divide and conquers. We found ParalTabs a useful algorithm to perform classification on genetic databases, obtaining improvements in execution times and performance measures. Using this system, we can determine genes associated with proliferation or non-proliferation of laryngeal cancer, only by taking samples of individuals and analyzing their DNA.

KEYWORDS:

• Classification
• Classifiers
• Data base
• Decision tables
• Genetic markers
• Laryngeal cancer
• Machine learning
• Parallel computing

ACKNOWLEDGMENTS

Special thanks to Consejo Nacional de Ciencia y Tecnologa – Conacyt and Universidad Autonoma Metropolitana – UAM for their support.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

¹ This paper is an extended version of two conference paper that appeared in [1] and [2]. The key additions of this journal version are as follows. First, in Section 3 we use a different way of how to implement the ParalTabs algorithm, using a set of processes instead of threads. Second, in Section 4 we use the same transformation of data used in previous work [2], but the size of data in this paper is different. Finally, this paper contains new experimental results in Section 4, one of this is using ParalTabs for searching genetic Q3 Q2 markers of laryngeal cancer, which is different from those shown in the paper [2].

Additional information

Notes on contributors

Benjamín Moreno-Montiel

Benjamín Moreno-Montiel received the BSc and MSc degrees in science and technologies of information from Universidad Autónoma Metropolitana – Unidad Iztapalapa, Mexico City, México, in 2007 and 2009, respectively. He is currently working towards the PhD degree at the Posgrado en Ciencias y Tecnologías de la Información in Universidad Autónoma Metropolitana – Unidad Iztapalapa, México City, México. His research interests are machine learning, ensembles of classifiers, and parallel computing.

E-mail: [email protected]

René MacKinney-Romero

René MacKinney-Romero received the BSc degree from Universidad Autónoma Metropolitana – Unidad Iztapalapa, Mexico City, México, in 1993, the MSc degree in computation from the University of Oxford, England, in 1994, and the doctorate degree in computer sciences from the University of Bristol, England, in 2002. He has been a professor of computer sciences since 1990 in Universidad Autónoma Metropolitana where he took his undergraduate courses in computer sciences. He teaches artificial intelligence and machine learning for undergrads and postgrads and has many publications in international conferences and journals.

E-mail: [email protected]

Carlos Hiram Moreno-Montiel

Carlos Hiram Moreno-Montiel received the BSc and the MSc degrees in science and technologies of information from Universidad Autónoma Metropolitana – Unidad Iztapalapa, Mexico City, México, in 2006 and 2010, respectively. He is currently working towards the PhD degree at the Posgrado en Ciencias y Tecnologías de la Información in Universidad Autónoma Metropolitana – Unidad Iztapalapa, Mexico City, México. His research interests are parallel computing, software engineering, and artificial Intelligence.

E-mail: [email protected]