A Novel Method for DNA Microarray Data Analysis: SDL Global Optimization Method - DNA Microarray Technology and Data Analysis in Cancer Research - page 40

two intensities of each dye is used as the gene expression data. The rela-
tive abundance of spotted DNA sequences in a pair of DNA or RNA
samples is assessed by evaluating the differential hybridization of the two
samples to the sequences on the array.
Gene expression levels can be determined for samples taken at multi-
ple time instants of a biological process or under various conditions. Each
gene corresponds to a high-dimensional row vector of its expression pro-
file (Mitra and Acharya, 2005). The goal of microarray experiments is to
identify genes that are differentially expressed in the conditions being
studied (Draghici, 2002). To compare gene expression profiles among
P190, P210, and P230, we expressed these three genes, respectively, in a
B-lymphoid cell line using a parental cell line that does not express BCR-
ABL as a control.

4.3. Fold Change Analysis

Fold change is defined as follows:

Take log 2 -transformed normalized intensities from RMA
(robust multi-chip averaging) for two samples;

Let a = log 2 (intensity sample #1)

Let b = log 2 (intensity sample #2)

If a > b : fold change = 2 (|a - b|)

If a < b : fold change = − 2 (|a - b|)

If a = b : fold change = 0.

Preprocessing of the gene expression profile is often necessary to
reach the goal of converting from raw data to biological significance. The
following steps are common:

• normalizing the hybridization intensities within a single array

experiment;

• transforming the data using a nonlinear function, like the logarithm in

case of expression ratios;

• estimating and replacing missing values in expressions, or adapting

existing algorithms to handle missing values;