The phrase ‘genome expression profile’ refers to the measured level of expression of a set of genes. Modern biotechnology allows scientists to measure the expression level of thousands of genes at once. This is done using what is called a microarray.
Suppose we are interested in how an increase in temperature affects the expression level of genes in a bacteria species. To quantify this using a microarray we maintain two groups of bacteria, one under a normal temperature and the other under an elevated temperature. Both groups of bacteria will generate various gene products and so, to tell them apart, we label all gene products from the ‘normal’ group with a green dye and all gene products from the elevated temperature group with a red dye.
The next step is to take copies of all known bacterial genes individually and adhere them to different locations on a glass plate. The dyed products from each of the two groups of bacteria are then washed over this plate, and each of the dyed gene products will bind to its corresponding gene. The result is a glass plate containing thousands of colored dots, one for each known gene. The dots will vary in color from red to yellow to green. A dot will be red if the corresponding gene tends to be expressed more under the elevated temperature than under the normal temperature. This is because most of the gene product bound to this gene on the glass plate will have come from the group of bacteria maintained under an elevated temperature (and this product has been dyed red). A dot will be yellow if the corresponding gene is expressed equally in both groups. Finally, a dot will be green if the gene tends to be expressed less under the elevated temperature. The entire glass plate is called a microarray.
To analyze a microarray it is first photographed. Then, a computer measures the intensity of red and green color for each dot in the image. If R and G denote the measured intensities of red and green respectively for a particular dot, then the ratio R/G gives a measure of the relative expression level of the gene in the two groups. For example, if R/G >1 then the expression level of this gene is higher in the group maintained at the elevated temperature. If R/G = 1 then the expression level of this gene is the same in the two groups. And if R/G < 1 then the expression level of this gene is lower in the group maintained at the elevated temperature.
Although the ratio R/G provides a natural measure of (relative) gene expression it has one important drawback. When a gene is upregulated (that is, when R/G > 1) the ratio can take any value in the interval (1,∞). On the other hand, when a gene is down regulated (that is, when R/G < 1), the ratio can take any value in the interval (0,1). Thus numerical scores for up- versus down-regulation are not comparable. For example, a gene expressed ten times as much under an elevated temperature would have a score of 10 whereas a gene expressed ten times as much under a normal temperature would have a score of 1/10.
To remove this asymmetry many scientists therefore work with the logarithm of the expression ratios. In this way, a gene that is upregulated will have a positive score and one that is down regulated will have a negative score. A gene that is equally expressed in both conditions will have a score of 0. Moreover, up- versus down-regulation are treated symmetrically. For example, working with logarithms to the base 10, a gene expressed ten times as much under an elevated temperature would have a score of 1 whereas a gene expressed ten times as much under a normal temperature would have a score of -1.