|
|
 |
/ Home / Analysis Tools / Images Analysis
DNA microarrays image analysis after microarray scan
Version 1.6
This tutorial can be download
as a PDF file (6 Mb).
The use of an image analysis software:
The ENS platform uses the GenePix Pro software. Its characteristics
are available on the Axon
internet website.
Software startup:
When GenePix Pro starts up, it is possible that some alert
messages appear. We will describe some of then, so that you know what to do
if they appear on screen:
- At the first software start-up, it can ask you in which colour mode you
want to work. It depends, of course, on the way you scan your slides. In our
case, we will select the “two colour” mode.
- The update window will pop out next. It is just necessary to click on the
“Continue” button.
- At start-up, GenePix Pro looks for scanners connected to the computer.
If the computer you are using is not connected to a scanner, or if the scanner
is switch off, select the “Analysis Only” mode:
- If not any scanner is detected; GenePix Pro will ask which scanner model
you wan to emulate. This is not so important, but it would have been preferable
to choose the scanner model you use to read your slides (use 4000B as default).
An introduction to the software interface:
Here is an introduction to the GenePix Pro software interface
(the screen shots come from the 4.1 and 5.1 software versions). A set of tabs
enables an easy navigation between the different software functions (image visualization,
result table, ...):

The image tab window is divided in 3 parts:
- the selection and the image navigation tools on the left
- the image visualization area in the middle
- a menu bar on the right
The toolbox on the left top of the window enables the control
of the image and grid settings:

Below this, an other tool enables you to visualize the different
wavelength images for the spot you point out with your mouse cursor, on the
microarray image. It also displays the spot intensity in each channel:

On the right part of the window, a menu bar gives access
to the scanner settings and the image analysis management. We will use these
following icons during the image analysis part:

Image acquisition:
It is necessary to do a first quick reading of the slide «
prescan » in order to locate the scan zone of the device.
It is important to pay attention to the gain and photomultiplicator
level (PMT) choice as to big values could increase artificially the electronic
noise of the device and sort out of the linearity range of the scanner. We advise
to read the slide using medium PMT level (around 600-650). It could be useful
to do several scans at various level of PMT in order to establish the PMT offset.
Generally, we use a pixel size of 10µm with only one
scan (lines to average = 1). However, setting this value to 2 or 4 reading allow
to reduce the electronic noise with a longer scanning time.
All along the scan, verify that the image is correctly equilibrated
in the “histogram” panel: the two curves corresponding to the two
channels must be superimposed for intensities greater than the background.
The different steps of a microarray image analysis:
Here are the practical steps to follow to achieve an image
analysis thanks to the GenePix Pro software:
- Open the image files with the file management menu icon
.The
image files are in tiff format. After scanning your microarray, you can save
your image as a single file (the compilation of the two wavelength images)
or as a multiple file (one file for each wavelength image). The multiple file
saving generates two files that are smaller than the single one and can be
opened in most of the image analysis software (the single file results from
the scanner software algorithm, and may not fit an other software to be opened).
Be careful, we will recommend to always save your images as multiple image
files, GenePix could sometimes have a very weird behaviour using single image
TIFF. If you open two TIFF images, you must select both of them at the same
time in the “Open Images” window.
- Now, it is useful to learn about a few buttons in the main GenePix window:
- First the contrast and brightness settings. These parameter modifications
enable you to have a better view of your spots on the microarray image.
You do not modify your images and raw data, but only what you see the
computer screen.
- You also have to know about the zoom mode that enables you to zoom in
and out to focus on a special block or spot. You can always get back to
your previous zoom level.
- After loading the images, you have to load the grid that references your
image: it is the GAL format file (for Gene Array List). It contains all the
informations that are needed to identify each spot (coordinates, name, identifiers)
on your microarray.
- Use the “Load Array List” function in the file management menu
to load the grid
.
Be careful to always save the files using a right clic on the link and using
the function “Save the target as…” from the contextual menu.


- Once it is loaded, you have to put it properly on you image.
- Choose the block mode
.
You can now select grid blocks and move them to combine them properly with
your image. The software is supposed to be able to do it automatically, but
in fact, you have to give it a little help. This step is the one you should
brighten and contrast the image enough to see the weak spots. Then, use the
spot at the bottom left of each block to position your grid: this spot is
the first one spotted on the slide. Be careful, pay attention to work in bloc
mode and not in the add bloc mode (the same icon but with a plus sign).
- Save your grid settings (use the function “Save settings” in
the file management menu). Your parameters are saved in a GPS file (GenePix
Settings). It is important to note that a GPS file is unique and dedicated
to specific microarray.
- Once your grid is positioned, use the “target” icon
to launch the automated identification of blocks. At the end of the process,
verify GenePix identification process has worked properly to avoid errors
such as mismatches on spot lines, block overlaps... To achieve this, move
a single block. Next, you can launch using the same icon the spot (feature)
identification inside each block.
- Do not forget to save your grid settings regularly.
- The next step constists in searching the spots that could be doubtful. GenePix
could have considered an artefact as a spot, rejected a spot you would have
accepted or mispositionned the circle around the spot. The more the image
quality is high, the less this step is long.
- Put yourself in spot (feature) mode
.
You can move the grid spots inside the block, blow them up or narrow them
to adapt the circle to the spots if you think they are not correct. You can
also flag them if you want to filter bad spot out. This avoids you to work
with spots that could be false positive such as spots with unbalanced fluorescence
due to dust.
- GenePix proposes different kinds of flags:
| |
|
|
| Absent spot (-75) |
Undetected spot (-50) |
Bad spot (-100) |
-
A right click on the spot (or a group of spots) when you
work in feature mode enables you to access a pop up menu with the spot parameters.
There, you can choose a flag to affect to the selected spot:
- At the end of the spot validation, launch the analysis
that is to say the conversion of pixels into digital intensity values.

- To get your result table, go to the result tab window and save them into
a file (GPR file for GenePix Results).
- GenePix may warn you that some spots do not have unique identifiers. In
our case, it is normal because some spots are registered as “empty”
in the GAL file (the grid file). Once you have saved your file, open it in
a spreadsheet program such as excel to supress them so that they can not be
taken into account in your analysis.

- You are now ready to normalize your data!
Useful links:
- GenePix® Pro is an acquisition and four-channel analysis software for
DNA and protein microarrays, tissue and cell arrays, and fluorescence studies.
Website: http://www.axon.com/gn_GenePixSoftware.html
|