Genepix Capture and Analysis (SOP #10)
Craig Story 1/9/07, modif 11/2/07
A. Genepix Capture.
(Tips and things to keep in mind.)
It is good idea to look over the manual from Axon Instruments to become familiar with terms such as feature, blocks, array, saturation. One can be found in drawer in microarray facility, and Chris has one that goes with the new 4 color scanner. It is much easier to have someone show you how to use the software for the first time.
Remember to place slides face down in scanner in Whitehead microarray facility (Axon Instruments Model 4000B). If you get a blank image, check this first. Slides should be rinsed PBS then in dH2O and spun-dry prior to scanning. Keep them sheltered from light. For the autoloader scanner in the Broad, be sure to get training to know how to set up the machine.
Use preview scan to adjust the gains appropriately for the wavelengths of interest. Try to get maximum brightness without saturation. Saturation is any pixel intensity greater than 65,335, and appears white. You should define a Scan Area around the area of interest on the slide before performing the actual scan.
Pixel size of 10 µm can be used for screening hybridomas, without line averaging. Using Line averaging of 2 or 3 and using 5 µm pixels will improve the quality of image by removing noise, but will lengthen time of scan (each line averaging doubles the time). These settings can be used with the autoloading scanner in Broad.
B. Genepix Analysis
- Open TIFF images with Genepix software
- Rotate to proper orientation
- Choose appropriate GAL file.
- Stretch/shrink, rotate to fit one of these into “cleanest” area of slide
- Copy and layout entire slide
- Do Feature Alignment
- Do data analysis on features
- Decide which spots are of interest and export these to Excel spreadsheet
1. Open TIFF images with Genepix software. (GenePix Pro 6.0) Files are usually in CCdom:gobo:ploegh_microscopy# (don’t open the jpg files)
2. Rotate to proper orientation. Use Image Tools button pull-down menu. 1. Flip left to right 2. Rotate left 90 deg. This should result in the upper right being #7 (on the 40x40 “new” style grids). Zoom in to make sure the notches are on the bottom and left edges of each sector.
3. Choose appropriate GAL file (Genepix Array List File). Use “save/open” button, and go to Load Array List [Alt+Y]. Navigate to ploegh_microscopy1:Clove:Microarray Data:GAL Files. Note: these can be Text files, make sure you are looking for “all files” You can also load a GAL file by going to “Open Settings” [Alt+O] and finding a recent GAL aligned file. Make sure the block layout is correct.
4. Stretch/shrink, rotate to fit one of these into “cleanest” area of slide. Use Zoom button, contrast/brightness, and “Block Mode” button to move block to best area of slide. Note, lower right corner can be used to expand and shrink entire block. The middle upper and left can be used to rotate it (pivots around upper left corner). Control-B causes the block overlay to appear and disappear. Make a new folder named “Analysis” and store all GAL and results files here.
5. Copy block and layout entire slide. You will need to have a full set of blocks laid out (so identity is correct for every one). They are automatically numbered from left to right up to down. In the case of 100 µm wells 40x40 there are 14 blocks. Be sure to save settings just before going forward with alignment [Alt+S].
6. Do Feature Alignment Use “Align Blocks” button (looks like yellow target) and modify the “options” as follows: Square features, resize min 70% max 150%, check limit feature movement <40 µm, flag features below threshold as “not found,” Set background threshold to 1000. uncheck warping and rotation. Hit “OK” once options are set. Then double check that you have saved the unaligned gal file [Alt+S] and proceed with alignment under the “Align Blocks” menu. Choose “Align Features in All Blocks” [Shift+F5]. If too many bad features are identified, modify options and redo alignment. When you have a good alignment, save settings as [Alt+H] “filename gals aligned”
7. Do data analysis on features [type F to be in feature mode, or click button to get out of Block mode].
This is the big moment we have all been waiting for: Click the “analyze button” (above the Save/Open button) [Alt+A]. Under “results” tab, data types, choose to display Block, Row, Column, ID, Flags, F532 Median –B532, F635 Median – B635, Log Ratio 635/532. You should also choose to not display “Not found” under “Display” in the Table section. [under “normalization” double check all except “not found”]
Next, use the Scatter Plot feature to quickly remove spots that are clearly artifacts, and identify/flag features. You can set X axis as 635, Y axis as 532, log scales. Often various types of artifacts will cluster together and can be lassoed and designated as “not found” or “bad.” Another useful way to graphically view the data is to change the Y axis to Ratio of Medians 635/532, or the Log of the ratio. Try various things! Once this is done, you can go back to image and make sure nothing obvious was missed. This is easy to do if the total number of spots is <100 or so.
Once your results are complete, you can save results [Alt+U].
There are several designations for each feature Vertical line = “not found” [N] (vast majority, do not display or list in results) Double Square = “good” [O] can be used to designate the “greens” X’d out = “bad” [A] Dashed line = “absent” [T] can use to designate the “reds” Open square = “unflagged” [L]
8. Export to Excel spreadsheet Export results as .txt file. This can be opened with Excel. Within Excel you can rearrange and rank them however you wish. Sorting by flag, and creating separate worksheets for each flag type is a good idea, also you may want to rank the ratio of colors as a measure of affinity.