Biology Reference
In-Depth Information
obtained from Student's
t
test. Because highly variable data can, by
chance, result in high or low fold changes, and data with little vari-
ance can result in low p-values despite very low changes, each test
alone can result in false positives (i.e., accepting biologically irrel-
evant data points as altered between mutant and wild type).
Therefore a combination of both criteria—fold change and
p-value—is recommended as it will minimize the false discovery
rate while maximizing the detection of reliable changes [
20
].
4
Notes
1. The collection of root and leaf material will allow comparing
the effect of a mutation on different Arabidopsis organs and
can be expanded to include other organs of interest. The prep-
aration of different organs is suitable to provide access to dif-
ferent sets of proteins because protein identification reflects
functional organ specialization. While leaf material is domi-
nated by photosynthetic proteins root proteins will cover other
functional categories. The difference in protein detection that
can be obtained by such a distinction is visualized in Fig.
2
(modified from ref. [
11
]). Root material was collected from
Fig. 2
Functional classification of protein identifications obtained from
Arabidopsis thaliana
leaves and roots
(data taken from ref. [
14
]). Categorization is based on the aspect “biological process” and the numerical values
are derived from p-values obtained with topGO using the
elim
method [
14
]. The Venn diagram shows the
number of protein identified from roots and leaves and those common to both organs
