Hi
I am using a Columbus for fluo images analysis. I need to define double or triple positive cells. My nuclei are stained with Hoestch (all cells), Sytox (dead cells only) and FITCs (apoptotic cells). I need to define which cells are double positive for FITCs and Sytox. Does anybody has any experience with that? Thanks!
CellProfiler [1] is the tool which should allow you to perform this study. It is too complex to explain here how you should use it, however after 2-3 introductory articles and after a week you should be able to crack the problem. Something on which you can try to build up is e.g. this topic [2]. Good luck!
[1] http://www.cellprofiler.org
[2] http://cellprofiler.org/forum/viewtopic.php?f=14&t=1560
I'm not familiar with Columbus, but if I'm correct you have 3 images of the same area (nuclei staining, dead cell staining (sytox), and apoptotic (FITC). For double staining you can just overlay all 3 the images right? I see that sytox is green, so is FITC, so you need to change the colour of one of the images (you can do that in photoshop for example), overlaying images can be done in Photoshop, or ImageJ. But a much easier way; what microscope did you use for the images? When I take pictures of cells that I stained with multiple dyes using the SP8 confocal microscope, the Leica software allows me to scan all 3 dyes (if you use two dyes in (approximately) the same wavelength you can do sequential scanning) and makes overlaying images. Double staining shows then up in yellow for me (since I use green and red colors).
I use Leica confocal microscope routinely and they have a software that helps me doing the pixel density/intensity count. For doing that, I need to open all the colored images at the same time, place a defined box such 20.20x20.20 on one of them, the box will generate the numbers and export them to the excel file. After taking pixel density of one image, when I switch to a different color it automatically places the box in to the same place that I had in the first one. Then it generates a different number and exports them to the excel. I usually take measurements from different regions of my sample and all those numbers get stored in to the database of the excel program. From where I generate the graphs for individual color.
Or you can work with either Photoshop and Image J as mentioned by both Dafna and Gopinathan.
Hope it helps. Good luck.
Hi, Thanks a lot everybody for answers and hints. I am using InCell HCS machine to take images (96 well format). They are acquired in tree different channels and it is intuitive to overlay the colors for co-localization. I think it is the Columbus analysis platform which has its specific ways to analyze the image. I keep on looking into it:)
I agree with Szymon, CellProfiler (Broad Institute) is probably the best way to go. We run 96well on INCell 2200 or Cytell on up to four channels and analyse with our custom-made CellProfiler pipelines. You could quickly modify one of the freely available CellProfiler pipelines or create you own. If you're planning on doing a lot image-based analysis, getting to know the power of the CellProfilers software modules is a worthwhile investment of time.
- Badges
- Science method
- Similar topics
- Biological Science
- Histology
- Staining