Hi Elizabeth,
An answer from a non-biologist, just considering the image-processing
aspects (I have no experience with western blots).
To determine an appropriate value of the radius, the best is to use
"create background" and "preview". Increase the radius until you see
no structure in the lanes; the lanes may remain weakly visible. If
the background of the original is rather even, you could increase the
radius a bit more. Then, switch off the "create background" option
for subtracting it.
There is not general rule for the radius as it depends on the scale
of your image and how much the the background varies.
If you don't subtract the background, the areas between the lanes
should have a roughly constant value. You could subtract this value.
Michael
________________________________________________________________
On 20 Nov 2008, at 18:19, Elizabeth Chia wrote:
> Hello all,
>
> I am just using ImageJ to analyze my western blots. I need help
> with the
> following questions:
> Process>Subtract background.
> (a)What is the appropriate value to use for the radius for a 8-bit
> image? I
> find that in some cases, the value 50 artificially bleaches parts
> of my
> image.
> (b)Do I need to subtract the background if the blot looks even?
> (c)If I don't use subtract the background, should I choose a random
> area of
> the blot as background?
>
> Thanks for your help in advance.
> Elizabeth
Locked
