Biology Reference
In-Depth Information
3.3. Software-Assisted
Image Analysis
After completion of the image acquisition process, the entire data
that constitutes an MFA experiment is exclusively contained in the
generated set of gel image fi les. It has to be emphasized that software-
assisted analysis of the image data can only provide meaningful
results, when it is conducted on the base of high-quality gel images.
Image analysis software is neither intended nor capable to attain
satisfactory (or even any) results from low-grade gel images.
The overall workfl ow for multifl uorescence 2D gel image analysis
is similar for all software packages available on the market:
-
Image cropping.
-
Joining gel images in groups of a MFA experiment.
-
Spot detection and matching.
-
Some important aspects of the analysis procedure are outlined below.
Difference analysis and extended statistical analyses.
Even when working with pre-defi ned scan areas, raw gel images
generated by the scan device contain non-essential information
that should be deleted prior to performing image analysis. This can
be realized by “cropping” these images, i.e. cutting the section of
interest from the raw image fi le and pasting it as a new image into
a new fi le. Within one MFA experiment, all cropped image sections
have to be exactly of the same size and should cover the same gel
section. Cropping can be performed using specialized software
that is capable of cropping all three corresponding images of one
gel in parallel without losing the calibrated image raw data. After
cropping, associated gel images can easily be linked to MFA experi-
ments of the established 2D image analysis software packages.
3.3.1. Cropping
A typical MFA experiment usually consists of a set of several gels.
After gel image acquisition, protein spots have to be detected and
all spots representing identical proteins within different gel images
have to be matched. For the images derived from the same gel,
spots do not need to be matched because of the co-migration of
differentially labelled proteins. However, considering all images
derived from different gels of one extensive MFA experiment, spot
matching becomes a challenging business and can justifi ably be
seen as the crucial step within the whole image analysis process.
Two different approaches for spot detection and spot matching
are currently realized: Most image analysis software packages such
as DeCyder™ (GE Healthcare) and PDQuest™ (Bio-Rad) use the
more “traditional” approach, whereas the Delta2D™ (Decodon)
software features an alternative spot matching concept.
The traditional approach starts with a spot detection routine to
detect protein spots in every gel image combined in the MFA
experiment. By using special spot detection algorithms in combination
3.3.2. Spot Matching
Search WWH ::
Custom Search