colocalization of the green and the red

65 views
Skip to first unread message

AMAR Majid

unread,
Jun 11, 2014, 9:56:02 AM6/11/14
to icy-so...@googlegroups.com

Hello,

I have one question about Icy,

 

I did a double staining (Red = CY3) (Green = FITC, Thioflavine)

So, I have a picture with
canal 1 = red

and canal 2 = green

I want to know if it’s possible in Icy to determine the overlapping surface (yellow) that corresponds to the colocalization of the green and the red, ( i.e. the yellow color)

Therefore, I want to quantify :

the surface of the red labeling

the surface of the green labeling

and the surface of the yellow (i.e. the colocalization of the green and the red)

 

could you please help me  with this?

 

Thank you

 

  Majid

AMAR Majid

unread,
Jun 12, 2014, 4:43:51 AM6/12/14
to icy-so...@googlegroups.com

To make the problem more simple,

I have one .zvi file

and (2 stainings: Red = Cy3 and Green = FITC)

I want to select an ROI (Region Of Interest)

And I would like some information regarding:

  1. The surface of the ROI (Number of pixels Total)

  2. The surface of the red labeling in this ROI (Number of Red Pixels)

  3. The surface of the Green Labeling in this ROI (Number of green Pixels)

  4. The overlapping surface in this ROI (number of yellow pixels) (i.e. number of red and green pixels, at the same time)

Thanks again

   Majid

Stephane

unread,
Jun 13, 2014, 4:59:41 AM6/13/14
to icy-so...@googlegroups.com
Hello Majid,

Actually that type of quantification is really easy to do in Icy :)
Just use some segmentation plugin and get result as ROI so you have direct quantification result in the ROI tab on the right.
For instance you can use the Thresholder plugin (in the "Detection & Tracking" tab, "Segmentation" section).
Sepcify the channel number where you want to perform the segmentation, specify the "K_MEANS" method with 2 classes (to separate object from background) and set "Output as" to ROI.
You should obtain a roi over the detected region.
Then you can repeat the operation with other channel.
Then you can process the intersection of both ROI to get overlapping region.
To process the intersection correctly you have to set the change the channel number of the ROIs to "ALL" as by default they use the channel number where you did the segmentation and so intersection of different channel will give you an empty result.
If you need to process many images the same way, you can use the "Protocols" plugin to design a workflow and apply it on a whole folder of images.
Hope that helps !

- Stephane

Igothigh

unread,
Jun 13, 2014, 8:24:28 AM6/13/14
to icy-so...@googlegroups.com
Hi Stéphane,

How can you change the channels from a specific one to ALL with protocol or script?

thanks.

Stéphane

Stephane

unread,
Jun 13, 2014, 9:37:43 AM6/13/14
to icy-so...@googlegroups.com

You have to use a script for that (or a javascript block if you use protocols) :

// retrieve selected roi of active sequence
r = getSequence().getSelectedROI()

// get roi 5D dimension
bounds = r.getBounds5D()
// set infinite C dimension
bounds.setInfiniteC()
// set back to roi
r.setBounds5D(bounds)



Le mercredi 11 juin 2014 15:56:02 UTC+2, AMAR Majid a écrit :

Curtis Rueden

unread,
Jun 13, 2014, 1:40:22 PM6/13/14
to icy-software
Hi Majid,

> I want to know if it’s possible in Icy to determine the overlapping
> surface (yellow) that corresponds to the colocalization of the green
> and the red, ( i.e. the yellow color)

Please be really careful when doing colocalization analysis. There are many pitfalls with doing it qualitatively. For details, see:


Regards,
Curtis


--
You received this message because you are subscribed to the Google Groups "Icy imaging" group.
To unsubscribe from this group and stop receiving emails from it, send an email to icy-software...@googlegroups.com.
To post to this group, send email to icy-so...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/icy-software/465fd6ab-db9d-440e-b3e0-399bc4550c87%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Reply all
Reply to author
Forward
0 new messages