her the microarray technology or the sequencing technology

s. The output of the former technology is called the microarray

ression data and the latter technology is called the sequencing

a. The phenomenon change can be observed in the output data of

chnology by comparing two conditions. The main objective of

ression pattern discovery is thus to search in the genetic signature

discover a small subset of genes which have significantly different

s to a stress across experimental conditions.

mportant understanding of a living system is that such a system

se all its genes to respond to a stress [Liu, et al., 2016; Antolovic,

19; Sidorenko, et al., 2019]. This is why the subset of genes,

ave significantly different responses to a stress across two

ntal conditions, is normally of a small size. To discover

ally expressed genes (DEGs) given two experimental conditions,

uired to compare genes one by one using a significance

ment. Such a measurement should be unified across different

nts or experimental data sets. Two commonly used significance

ments are the biological significance and the statistical

nce [Chagoyen and Pazos, 2010; Minguez and Dopazo, 2011;

n, et al., 2017; Miller, et al., 2019].

e biological significance

ogical significance used to discover DEGs is the so-called fold

r the base two logarithm fold change. It has been exercised in

ince 1960s [Pauly, 1960; Hill and Sussman, 1964; Gilbert and

n, 1969]. Suppose a case condition is denoted by A and a control

is denoted by B. The case condition normally represents the

ntal group in which a stress is applied or a phenomenon change

observed. The control condition normally represents the group in

stress is applied or no phenomenon change has been observed.

change shown below is defined as the ratio between two mean

ns (ߤ௚,஺ and ߤ௚,஻). Note that ߤ௚,஺ and ߤ௚,஻ stand for the mean

ns of the g^th gene.