To search, Click below search items.


All Published Papers Search Service


A hybrid genetic algorithm and artificial immune system for informative gene selection


Mohammed Korayem, Waleed Abo Hamad, Khaled Mostafa


Vol. 10  No. 7  pp. 76-83


In this paper, we present a general approach for gene selection of high dimensional DNA Microarray data. The proposed approach represents a powerful new tool in the analysis and exploration of complex data. Very few genes are assumed to anticipate the pathological behavior of cancers. To this end, we proposed a hybrid between genetic algorithms and artificial immune system method it takes into account the main immune aspects selection and cloning of the most stimulated cells, death of non-stimulated cells, affinity maturation and reselection of the clones with higher affinity, generation and maintenance of diversity, hypermutation proportional to the cell affinity. The proposed approach is experimentally evaluated on the widely studied Colon, Leukemia and Lymphoma data sets. The results show that our approach is able to obtain very high classification accuracy which emphasizes the effectiveness of the selected genes and its ability of filtering the data from irrelevant genes. Also the criterion of the number of genes was integrated into the fitness function. Obtaining multimodal solutions is a major strength point of our method, only biologists and medical scientists can say which one of these solutions (gene subsets) is more biologically relevant to cancer diagnosis.


Artificial immune system, Genetic algorithms, DNA microarray Data, Classification, Gene Selection