distance based co-localization as bug as macro in JACoP?

[Stephen Taylor]

Hi,

I am emailing regarding what looks to be an issue with the JACoP plugin when using distance based co-localization as a macro.

When I run the JACoP interactively and set macro to record from imageJ (version 1.45s) on Windows 7, the macro code looks like this

run("JACoP ", "imga=[ROI_0 (red)] imgb=[ROI_0 (green)] thra=128 thrb=128 objdist=4-6650-212.62857142857143-497.95918367346945-true-false-true objcentpart=4-6650-true-false-true");

and on clicking 'Analyze' it shows only foci that are only colocalizing foci based on the distance metrics. When running the above code in a macro I get an image that seems to be showing *all* foci. Ideally I would like only colocalizing foci to be shown.

The log produced during macro/batch mode also has this error, so the number of colocalizing foci is *always* the same as total number of foci with this method. An example of part of the output is:

===========================================================
Colocalization based on distance between centres of mass
Threshold for Image A=128; Image B=128
Particles size between 1 & 9030
Image A: 28 centre(s) colocalizing out of 28
Image B: 9 centre(s) colocalizing out of 9
============================================================

I downloaded the plugin from http://imagejdocu.tudor.lu/doku.php?id=plugin:analysis:jacop_2.0:just_another_colocalization_plugin:start

Interestingly, there seems to be similar sounding bug that has been fixed on the wiki page

2.1.1 (20/08/2010):

    Fixed a bug about distance based co-localization when calling the function from a macro.

Is the 2.1.1 version available?

Thanks for any help,

Steve

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Stephen Taylor]

On further investigation of the object based analysis distance calculation there seems to be a bug in the d(A-B) calculation.

In the report the column d(A-B) is not taking into acount xy calibration, which is 67nm in the example below. Instead it seems to assume it is 1, so when it compares to the reference_dist everything passes (since it is much lower) hence everything appears colocalised. For example:

Interactive JACoP
        Label Centre_A_n° Centre_B_n° d(A-B) reference_dist phi theta XA YA ZA XB YB ZB
548 Colocalization 19 26 0 497.959 1.571 0 65.5 50 0 65.5 50 0
66 Colocalization 3 8 13.421 212.629 0.588 1.571 32.333 20.889 0 32.5 21 0
579 Colocalization 20 28 14.357 212.629 0 1.571 88.286 57 0 88.5 57 0
397 Colocalization 14 20 42.375 212.629 1.893 1.571 17.2 41.6 0 17 41 0
399 Colocalization 14 22 94.752 212.629 1.713 1.571 17.2 41.6 0 17 43 0
98 No colocalization 4 11 243.884 212.629 0.278 1.571 91.5 22 0 95 23 0
35 No colocalization 2 6 247.185 212.629 2.82 1.571 86.5 20.167 0 83 19 0
122 No colocalization 5 6 276.248 212.629 1.326 1.571 82 23 0 83 19 0
338 No colocalization 12 19 284.257 212.629 0.785 1.571 45 38 0 48 35 0

Macro/Batch JACoP
        Label Centre_A_n° Centre_B_n° d(A-B) reference_dist phi theta XA YA ZA XB YB ZB
548 Colocalization 19 26 0 497 1.571 0 65.5 50 0 65.5 50 0
66 Colocalization 3 8 0.2 212 0.588 1.571 32.333 20.889 0 32.5 21 0
579 Colocalization 20 28 0.214 212 0 1.571 88.286 57 0 88.5 57 0
397 Colocalization 14 20 0.632 212 1.893 1.571 17.2 41.6 0 17 41 0
399 Colocalization 14 22 1.414 212 1.713 1.571 17.2 41.6 0 17 43 0
98 Colocalization 4 11 3.64 212 0.278 1.571 91.5 22 0 95 23 0
35 Colocalization 2 6 3.689 212 2.82 1.571 86.5 20.167 0 83 19 0
122 Colocalization 5 6 4.123 212 1.326 1.571 82 23 0 83 19 0
338 Colocalization 12 19 4.243 212 0.785 1.571 45 38 0 48 35 0

This looks like it should be easy to fix, is the latest source code available for the plugin?

Thanks,

Steve

>
> Hi,
>
> I am emailing regarding what looks to be an issue with the JACoP plugin when
> using distance based co-localization as a macro.
>
> When I run the JACoP interactively and set macro to record from imageJ
> (version 1.45s) on Windows 7, the macro code looks like this
>
> run("JACoP ", "imga=[ROI_0 (red)] imgb=[ROI_0 (green)] thra=128 thrb=128
> objdist=4-6650-212.62857142857143-497.95918367346945-true-false-true
> objcentpart=4-6650-true-false-true");
>
> and on clicking 'Analyze' it shows only foci that are only colocalizing foci
> based on the distance metrics. When running the above code in a macro I get
> an image that seems to be showing *all* foci. Ideally I would like only
> colocalizing foci to be shown.
>
> The log produced during macro/batch mode also has this error, so the number
> of colocalizing foci is *always* the same as total number of foci with this
> method. An example of part of the output is:
>
> ===========================================================
> Colocalization based on distance between centres of mass Threshold for Image
> A=128; Image B=128 Particles size between 1 & 9030 Image A: 28 centre(s)
> colocalizing out of 28 Image B: 9 centre(s) colocalizing out of 9
> ============================================================
>
> I downloaded the plugin from
> http://imagejdocu.tudor.lu/doku.php?id=plugin:analysis:jacop_2.0:just_anothe
> r_colocalization_plugin:start
>
> Interestingly, there seems to be similar sounding bug that has been fixed on
> the wiki page
>
> 2.1.1 (20/08/2010):
>
>     Fixed a bug about distance based co-localization when calling the
> function from a macro.
>
> Is the 2.1.1 version available?
>
> Thanks for any help,
>
> Steve

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[tibeon]

Hi,

I have exactly this problem with a macro using the JACoP pluging. Did you find any solution since your last posts below ?

thanks a lot
benoit theze