2D DIGE Analysis of Protein Extracts from Muscle Tissue - Difference Gel Electrophoresis (DIGE): Methods and Protocols - page 163

3.8. Image Analysis

Various fully automated DIGE image analysis software packages
(DeCyder from GE Healthcare, Progenesis SameSpots from
Nonlinear Dynamics, Melanie from Bio-Rad) developed for spot
detection, matching, and differential protein expression analysis
are commercially available.
The most important steps involved in image analysis are:

- Spot detection.

- Background subtraction.

- In-gel normalization.

- Artifact removal.

- Gel matching.

- Statistical analysis.

Here, we describe the principal steps of image analysis per-
formed using the DeCyder software ( 5, 8, 11-13 ).

The software consists of three main modules:

DIA (differential in-gel analysis): to perform protein spot detec-
tion and quantifi cation on a pair of images, from the same gel.
A set of images is merged together, thereby incorporating all spot
features in a single image. Spot detection and spot boundary defi -
nition is performed using pixel data from all the individual raw
images and the merged image. An estimation of the number of
spots present on the images must be entered—it is recommended
that this value is overestimated to compensate for the detection of
nonproteinaceous spots on the image. One can later use the
Exclude fi lter in order to eliminate nonproteinaceous spots (based
on their physical characteristics). Numerical data for individual
spots are calculated (volume, area, peak height, and slope). Spot
volumes are always expressed with the background subtracted.
Background is subtracted by excluding the lowest tenth percentile
pixel value on the spot boundary from all other pixel values within
the spot boundary. The spot volume is the summation of these
corrected values.
BVA (biological variation analysis): to match multiple images from
different gels to provide statistical data on differential protein
expression levels between multiple groups.

EDA module (extended data analysis): to elaborate a set of data.
There are different calculations that can be performed in EDA:

- Differential Expression Analysis . This kind of calculation per-

mits to investigate differential expression between two or more
experimental groups. There are different subanalyses related to
the experimental setup. Student's t test is used to evaluate the
hypothesis of a difference between two groups. If more than
two groups are included into the experimental design, ANOVA
must be used. The simplest form of ANOVA is known as