Protocols: Quantification with ArrayVision
This protocol describes how to use ArrayVision (Imaging Research,
Inc.) to quantify images such as those produces by the MD Anderson
Cancer Center Genomics Core Laboratory.
This protocol makes the following assumptions:
- The scanning protocol was followed .
- The design has 4 rows and 12 columns of patches, or grids. Each
grid is 10x10 spots. Note that the images produced are typically
rotated 270 degrees (clockwise) from the slide design, so that there
are 12 rows and 4 columns.
- Run ArrayVision
- The most likely error message is that too many copies of
ArrayVision are already running. We have purchased a limited number
of network licenses, which restricts the number of people who can
run the program simultaneously. If this occurs, wait and try again
- When the program starts, you should see a window offering various
wizard options. Make sure "Analysis wizard" is selcted, and then
press the "OK" button.
- Press the "Next" button on the window labelled "Welcome to the
- The next window allows you to choose the analysis protocol. The
preferred protocol is called "ACCG 2-color 4800 array". If this is
available, select it from the drop-down list near the top of the
window, and then press the "Next" button.
- Note: If the preferred protocol is not available, you must
either import it from another location on the network (in which case
you should ask for help), or you must create your own by following
the protocol for creating ArrayVision
- The next window allows you to load the images. You can only use
it if no preprocessing (such as rotating or cropping) is needed
before loading the images. By default, the wizard opens a window
that is built from a standard file-selection tool, and will allow
you to load two images at once. It will almost certainly start in
the wrong directory, so you'll have to maneuver around a to locate
the correct directory.
- If you must rotate or crop the image in order to use it,
close the file selection window, and then press the "Images" button
on the remaining window. Make the selection "Retrieve image into
channel". (Do not choose "Retrieve images for all channels".) This
window includes an "Options" button that allows you to specify
rotation and cropping information. Use this procedure to load the
images one at a time. The options set on the first image will be
remembered for the second image.
- If the images were produced following our scanning protocol,
then they should have names of the form "blahblahA.TIF" or
"blahblahB.TIF". So, you will have to change the file type in the
"Select Images" window to "TIFF". (Do not use the similar selection
"TIFF 5", which refers to an older version of the file format.)
Start by loading the "A" image into channel 1; this represents the
red, Cy5-labelled sample. Next, load the "B" image into channel 2;
this represents the green, Cy3-labelled sample. Finally, press the
- This will return you to the "Select Images" portion of the
analysis wizard. Verify from the display that you have loaded the
correct images into the correct channels, and then press the "Next"
- You can make various adjustments to the displayed image to make
the spots easier to see. (These adjustments do not change the values
that are used for quantification.) The simplest adjustment is to
press the auto-contrast button (a circle with eight rays coming
out). Some slides are improved by choosing "Logarithmic" from the
drop-down box that initially contains the "Linear" choice.
- Verify that the alignment window includes the "Align all
channels" label, indicating that both images will be aligned
together. If this label does not appear on a gray background, return
to the protocol-selection step and make sure you have selected the
- Press the "Position" button and follow the instructions. This
step merely sets the upper-left-hand corner of the template. You
must make sure the "Alignment" window does not cover the upper left
corner of the grid in the image window before pressing
"Position". When the left-corner is placed, press "OK".
- Adjust the template so it matches the actual layout of spots
reasonably well. The handle at the upper left corner of the template
allows you to reposition it. The handle at the upper-right allows
you to rotate. Grabbing and dragging any side allows you to change
the size. By clicking to the left (or right) of any row of grids,
you can move, rotate, or resize the part of the template for that
row. In a similar fashion, you can click above (or below) to select
any column of grids. You can select a single 10x10 grid by clicking
anywhere inside it.
- When you are finished with the rough adjustment of the subgrids,
click above and to the left of the entire template to ensure that
the whole array is selected. Then press the "Align" button to allow
the software to refine the positions of individual spots.
- Inspect the alignment visually, and repeat until you are
satisfied that the template is correctly aligned with the spots in
the image. Then press the "Next" button.
- This opens the "Complete analysis" window, which actually
generates the numbers. Make certain that "Perform sampling and
post-analysis operations" is selected, then press the "Finish"
- After sufficient time passes, ArrayVision will open a "Save Data
File" window. Make sure this is pointing to the directory where you
want the "lg2" file saved, give it a meaningful name, and press
"Save". Do the same thing when asked to save a ".txt" file. (The
first file is ArrayVision's native format, and it will allow you to
review the alignment if questions arise about it in the future. The
second file is an ASCII tab-separated-value file that is suitable
for inclusion into other analysis programs, and is the preferred
format for uploading into the ACCG database.
- Proceed to the data upload