Thank you and to people like you that put the effort into making these tutorials. Is not easy to find nice and explicative tutorials, like yours, in these kind of programs.
Hello! Thank you for sharing. I have a question. In the case you r working with samples such as a liquid one, and you can distinguish standard fluorescence fruom structure as you did with cells. I think ACE will do the job for you executing the clusterinzing algorithm. Is it not a secure way of components extraction? Best Regards from UNICAMP, sp, brazil
hejsan!sorry, i am not sure, which of the ACEs do you have in mind. adaptive clustering algo? i haven’t used much of the custom methods for automated image segmentation. You might be interested in some of the already available pipelines for it in MatLab and Ilastic
without choosing the point where to extract the ACE component. cause, i work with a solution of fluorescent molecules all mixed and i cant choose the component by choosing the area at the image. so, for my case, the clustering algorithm is the only option. I prefere to treat data direct on/in the microscope, i use ZEN. My point is: To automate extraction is not reliable?
@@manoellourenco000 I have never used ACE myself. From Zeiss description it looks not extremely accurate (Figure 4d): zeiss-campus.magnet.fsu.edu/articles/spectralimaging/considerations.html
Thanks a lot Alyona for making this video. This is very useful and informative. I have trying to image amcyan, fitc and mcherry in my samples. However, I think there is a cross talk between amcyan and fitc. Even if I make an individual lambda stack for both the images, there would be still an overlap. If i try to shorten the spectral scan range manually, I might loose signal information. At the moment, I am using best signal for scanning my flourophores. If I scan all 3 flourophores in lambda mode, can i spectrally separate them later? I need to perform co-localization studies with amcyan and fitc and amcyan and mcherry, however I must first make sure that there is no overlap/leak which might give a false positive result. I can send you my zen file, if you want to have a look. Could you also please do a video for quantitative co-localization measurements like manners coefficient.? thanks again
Yes, you can spectrally separate your lambda stack images after scanning, but it might lead to the same aberrations as narrowing down the detection range. Generally Linear unmixing will change intensities values on your image and colocalization calculation will be based on comparing intensities of different channels for the same pixel positions, so one should unmix with extreme care in order not to introduce artifacts and only if there is no better fix for the situation. please try to check for a crosstalk by switching off excitation lasers one by one and verifying that the corresponding channels disappear completely. I show how to do it in the FRET video from 5:00 ua-cam.com/video/CODwQO6TUB4/v-deo.html You might be able to tune laser intensity/Master gain/detection range to bring the crosstalk to a neglegtable level. AmCyan and FITC have significant overlap in excitation and emission. BD biosiences claim that BD Horizon V500 leaks into FITC less ( I haven't tried it) If you can scan in vivo and your labeled structures move, it is very useful to make movies and then track "co-movement" of different fluorophores I have colocalization video on my list, but I can't promise to make it soon. I am a full-time researcher, running confocal platform at the BioC comes on the top of it, so time is a big problem. There is a cool plug in ImageJ, which will calculate for you Manders and Pearson coefficients imagej.net/Coloc_2. I'm sure you'll be able to find also a plugin for Spearman's colocalziation coefficient just by googling it. May I ask you, where are you located? Do you get any support for the microscopy related issues?
Thank you and to people like you that put the effort into making these tutorials. Is not easy to find nice and explicative tutorials, like yours, in these kind of programs.
Thank you for your help with this tutorial.
Hello, why is there no difference in fluorescence intensity at different emission wavelengths in my lambda scan?
could you share your image file?
Thank you! It's very helpful.
Hello! Thank you for sharing. I have a question. In the case you r working with samples such as a liquid one, and you can distinguish standard fluorescence fruom structure as you did with cells. I think ACE will do the job for you executing the clusterinzing algorithm. Is it not a secure way of components extraction? Best Regards from UNICAMP, sp, brazil
hejsan!sorry, i am not sure, which of the ACEs do you have in mind. adaptive clustering algo? i haven’t used much of the custom methods for automated image segmentation. You might be interested in some of the already available pipelines for it in MatLab and Ilastic
@@AlyonaMinina The ideia is to extracte ACE from Lambda Mode aquisition
without choosing the point where to extract the ACE component. cause, i work with a solution of fluorescent molecules all mixed and i cant choose the component by choosing the area at the image. so, for my case, the clustering algorithm is the only option. I prefere to treat data direct on/in the microscope, i use ZEN. My point is: To automate extraction is not reliable?
@@manoellourenco000 I have never used ACE myself. From Zeiss description it looks not extremely accurate (Figure 4d): zeiss-campus.magnet.fsu.edu/articles/spectralimaging/considerations.html
Thanks a lot Alyona for making this video. This is very useful and informative.
I have trying to image amcyan, fitc and mcherry in my samples. However, I think there is a cross talk between amcyan and fitc. Even if I make an individual lambda stack for both the images, there would be still an overlap. If i try to shorten the spectral scan range manually, I might loose signal information. At the moment, I am using best signal for scanning my flourophores. If I scan all 3 flourophores in lambda mode, can i spectrally separate them later?
I need to perform co-localization studies with amcyan and fitc and amcyan and mcherry, however I must first make sure that there is no overlap/leak which might give a false positive result. I can send you my zen file, if you want to have a look.
Could you also please do a video for quantitative co-localization measurements like manners coefficient.?
thanks again
Yes, you can spectrally separate your lambda stack images after scanning, but it might lead to the same aberrations as narrowing down the detection range. Generally Linear unmixing will change intensities values on your image and colocalization calculation will be based on comparing intensities of different channels for the same pixel positions, so one should unmix with extreme care in order not to introduce artifacts and only if there is no better fix for the situation.
please try to check for a crosstalk by switching off excitation lasers one by one and verifying that the corresponding channels disappear completely. I show how to do it in the FRET video from 5:00 ua-cam.com/video/CODwQO6TUB4/v-deo.html
You might be able to tune laser intensity/Master gain/detection range to bring the crosstalk to a neglegtable level.
AmCyan and FITC have significant overlap in excitation and emission. BD biosiences claim that BD Horizon V500 leaks into FITC less ( I haven't tried it)
If you can scan in vivo and your labeled structures move, it is very useful to make movies and then track "co-movement" of different fluorophores
I have colocalization video on my list, but I can't promise to make it soon. I am a full-time researcher, running confocal platform at the BioC comes on the top of it, so time is a big problem. There is a cool plug in ImageJ, which will calculate for you Manders and Pearson coefficients imagej.net/Coloc_2. I'm sure you'll be able to find also a plugin for Spearman's colocalziation coefficient just by googling it.
May I ask you, where are you located? Do you get any support for the microscopy related issues?