Login  Register

Re: Adjust resolution of "plot lanes" command

Posted by Luke Goodsell on Feb 24, 2010; 5:34pm
URL: http://imagej.273.s1.nabble.com/Adjust-resolution-of-plot-lanes-command-tp3689205p3689208.html

Thanks for the response. This comes close to achieving what I'm after.
After setting it to vertical, with no grid lines and inverting my
image, I can see the kind of graph I'm after. However, when I try to
close a peak with the line tool and then use the wand tool to select
the area under the peak, the line I drew disappears and the wand tool
selects the entire plot (as in the attached image. Is it possible to
manually close a peak and then have ImageJ tell me the bound area?

Many thanks,
Luke

On Wed, Feb 24, 2010 at 4:21 PM, Gluender <[hidden email]> wrote:

> Luke,
>
> it's much clearer now: you're concerned about the amplitude (gray-value).
>
> I'm not familiar with the built in gel-analysis but it should be based on
> profile plots. That said, I recommend to do the following:
>
> 1.
> Go to menu item "Edit > Options > Profile Plot Options" set the desired
> window dimensions of the plot window. In your case set Height to perhaps
> your screen height or even higher.
>
> 2.
> Make a suitable rectangular selection for the lane of interest.
>
> 3.
> Go to menu item "Analyze > Plot Profile"
>
> Is the plot you get better?
>
> Now you may even have a look at the amplitude values, if you click "List" in
> the plot window.
>
>> Hi Herbie, thanks for your response.
>>
>> The 'resolution' I'm referring to is that of the graph images produced
>> by the 'plot lanes' operation. The width is the same as that of
>> selection height on my image, which is fine, but the height is a
>> constant value with the height of the peaks scaled according to the
>> highest peak. However, I'm using chemiluminescence and a 16bit tif, so
>> my peak heights can vary considerably. When I try to quantify a small
>> peak, I cannot accurately draw the baseline on the peak as it is only
>> a couple of pixels high. Zooming in on the graph just magnifies the
>> pixels of the graph, offering no increase in accuracy.
>>
>> I would like to be able to increase the (seemingly hard-coded) height
>> of the graphs produced to be more useful for 16 bit images. Is this
>> possible?
>>
>> Many thanks,
>> Luke
>>
>> On Wed, Feb 24, 2010 at 2:47 PM, Gluender <[hidden email]> wrote:
>>>
>>>  Luke,
>>>
>>>  not quite sure what the real problem is...
>>>
>>>  First, there is the spatial resolution of your images.
>>>  This resolution will determine most of your post-processing. If it is to
>>>  low, then you have to increase it by scanning/capturing parts.
>>>  Do you scan the lanes or do you use a camera?
>>>
>>>  Second, there is the spatial resolution of your profile graphs (x-axis).
>>>  This resolution at best can be the same as the spatial resolution of
>>> your
>>>  images.
>>>  It may however be smaller. You may change it in Menu-Item "Profile Plot
>>>  Options".
>>>
>>>  Finally there is the amplitude (gray-value) resolution of your images as
>>>  well as your profile plots (y-axis). 8 bits means 255 values different
>>> zero
>>>  and under all circumstances your profile graph can't show more. Plot
>>>  amplitude resolution may also be changed in Menu-Item "Profile Plot
>>>  Options".
>>>
>>>  That said you may give us a little more information about your desires.
>>>
>>>>  Hi,
>>>>
>>>>  I am trying to quantify my westerns but am having difficulty because
>>>>  the resolution of the graphs output by the "plot lanes" command is too
>>>>  low. I have a number of bands of a wide range of concentrations so the
>>>>  smallest bands have small peaks in the graphs, making it impossible to
>>>>  place the baseline accurately. I assumed the zoom tool would help, but
>>>>  that just, err, makes the pixels bigger - sorry, I don't know how to
>>>>  describe it. I also tried using multiple selections in the same lane,
>>>>  but then the areas of the graphs are not comparable because they have
>>>>  been scaled differently. I've attached an example of the graphs I get.
>>>>
>>>>  I searched Google and the mailing list archives, and I saw at least
>>>>  one person mention this as a side issue, but no one addressed it.
>>>>
>>>>  Is it possible to change the resolution/size of the graphs output by
>>>>  the "plot lanes" command so that I can zoom in and accurately define
>>>>  the baseline?
>>>>
>>>>  Many thanks,
>>>>  Luke
>>>>
>>>>  Content-Type: image/png; name="Picture 1.png"
>>>>  Content-Disposition: attachment; filename="Picture 1.png"
>>>>  X-Attachment-Id: f_g627gugk0
>>>>
>>>>  Attachment converted: Download:Picture 1 3.png (PNGf/«IC») (0001832A)
>>>
>>>  Best
>>>  --
>>>
>>>                  Herbie
>>>
>>>         ------------------------
>>>         <http://www.gluender.de>
>>>
>
> HTH
> --
>
>                  Herbie
>
>         ------------------------
>         <http://www.gluender.de>
>