K-Means Clustering on 3rd order polynomial based normalization of Acute Myeloid Leukemia (AML) and Acute Lymphocyte Leukemia (ALL)

Microarray expression data is one of the most widely used to find patterns in genetic expressions. The DNA microarray technique participates as one of the leading methods in cancer research. Due to the presence of immense noise, fewer numbers of samples and huge amount of genes, the useful genomic knowledge extraction from this technique is an important question in today's biological research. Scientists and researchers are exploring efficient mathematical procedure to find realistic gene expressed knowledge. In this study K-Means clustering technique is used on an efficient 3rd order polynomial based technique to normalize the genomic data of acute myeloid leukemia (AML) and acute lymphocyte leukemia (ALL). AML was used as a model to generate the coefficients of the polynomial by considering non trending, decorellation and offset based techniques. The K nearest neighbor technique is used to estimate the missing values of microarray data and avoid the impact of missing data on clustering algorithm. The data can be regenerated easily using 3rd order polynomial normalization based on model generated by AML. Top ranked genes in each cluster have been presented in this paper which helps in finding functionally coregulated genes in ALL and AML. Proceedings of a meeting held 9-11 April 2009, Lahore, Pakista