Instructions for Processing Images

This document is also available for download in pdf or word format.


After scanning the mArray slides, one should have two images, green (532) and red (635) for each slide.  We put red and green fluorescent markers on a slide to double the amount of information we can gain from an experiment so we use the numerical information from each color during the analysis.  To process the images for viewing, however, we need to merge the two colors into one image.

Analyzing the images is time intensive so we want to view the image of each slide to get an overall gestalt for the experiment and evaluate how good or bad the experiment is before we go through the trouble of analyzing.  When evaluating a slide, look for how high the background intensity is, whether the spots are uniform morphology, whether the control markers have uniform intensity, and whether the slide looks beautiful.  A clean and aesthetically appealing slide usually means a good experiment.

Notes on ScanAlyze

Michael Eisen developed the ScanAlyze program when he was at Stanford University.  The purpose of this program is to “Process fluorescent images of microarrays.”  The ScanAlyze program is only available for Windows.  For more information, check out: Eisen’s homepage includes the manual, software, and links to other sites)

ScanAlyzeDoc.pdf (Manual for ScanAlyze)


Two Become One

Launch ScanAlyze 2.50

Select Load Channel 1 (button against the green backdrop)

Select the slide with the green (532) extension

Select Load Channel 2 (button against the red backdrop)

Select the same slide as above with the red (635) extension

Adjust the Gain to 7.5 (this value was empirically derived and gives the processed image an optimal intensity)

Select Redraw

Select Save

Name the merged image.  It’s a good idea to include the slide number, print date, and gain setting for the image.  The merged image now has a *.bmp extension.

Continue merging the two colors of each slide until all the slides are done.

Six Images on a Page

Launch Adobe Illustrator 10.0.3

Select File, New (Ctrl+N)

Select OK to chose the default document

Select the Type Tool (T)

Select somewhere near the top of the document and type in some information that is pertinent to the images on the document.  This information could include: the experiment name/number/date, print date, slide numbers, and gain setting.  For example, here’s the format I use: 

    Experiment XX, Human Sample Y    slides XX – XX    print MM/DD/YY    probe MM/DD/YY    gain 7.5

Select File, Open (Ctrl+O)

Select the first image from the experiment

Copy the image (either right click on the image and select copy or press Ctrl-C)

Close the image

Paste the image onto the document

Press the Right mouse button somewhere on the image and select Transform, Scale…

Adjust the Uniform scale to 12% and select OK

Move the image to a position in the upper left area of the document.

Repeat the steps for adding an image to the document for the next five images.

Select the Type Tool (T)

Select somewhere near the top of each image and type in some information that is pertinent to the particular slide such as, slide number and sample name.

Select between the two columns of images and type in information that is pertinent to the experiment such as, information on the primary sample and secondary antibody.

In the end, one should have a pattern as shown below.

Select all six images by clicking on each while holding the shift key.

Select Effect, Rasterize…

Select OK to chose the default settings

Select File, Save As… (Ctrl+Shift+S)

It’s a good idea to include the print data and the number of all slides that appear on the document (X – X is fine). The document now has a *.ai extension.

Select File, Print…(Ctrl+P)

It’s a good idea to print a hard copy for the notebook or other records.  Plus, when all the slides are done one can look at the complete collection to get a feel for the quality of the experiment.

Continue making documents with six slides on a page until all the slides from the experiment are finished.

Brian A. Kidd Ó 2004