International Journal of Computational Intelligence, Biotechnology and Biochemical Engineering

Authors:- L Nishitha, Manasa S K, Manisha S K, Nithya A, Meenakshi

Abstract:-Microarray is one of the most promising tools available for life sciences researchers to study about gene expression profiles. Gene expression levels can be obtained through microarray analysis and the biological information of a disease can be identified. Cancer is a group of diseases which involves uncontrollable cell division with the potential to spread to other parts of the body destroying neighboring organs causing bleeding, blockages and loss of production of normal biochemical products. It results in decreased immunity and loss of body weight. Lung cancer is one of the leading cancer causes of death in this highly progressive world. An abysmally low rate of cure shows the inclination of lung cancer to usually be present as clinically advanced tumors. Most lung cancers are discovered very late, by which time the option for effective treatment is very limited. It is very important that the general physicians recognize the tumor or a lesion for what it is and know how to proceed with the investigation. Lung cancer can be identified at a time that is far more curable by predicting the chances of an individual getting said cancer. This can be done by matching an individual’s DNA with the diseased DNA (DNA affected with lung cancer) and matching them to calculate the chances of the individual being affected with cancer. For this, subset of genes is selected from a large dataset with redundant genes (noisy data) so that their features are distinguishable between normal and diseased gene samples. In this paper, k-nearest neighbor (k-NN) method is used to perform sample classification of genes and make use of a meta-heuristic algorithm for optimal matching in order to predict an individual’s chances of getting lung cancer.