quantifying PCR bands

Irit Rappley via methods%40net.bio.net (by irappley from scripps.edu)
Thu Feb 17 23:46:51 EST 2011


I would use ImageJ, not Photoshop.

The best quantification method is integration. Let's say that you draw a 
rectangle around each band, then calculate the pixel intensity of each 
column of pixels and make a histogram of these values, with pixel 
location on the x-axis and intensity value on the y-axis. The amount of 
signal in the band is represented by the area under the curve -- the 
integral.  Photoshop can't calculate this, but I think there might be a 
plug-in for ImageJ that can.

So what you would do is this: draw a rectangle around each band. 
Calculate the integrated intensity for each band. Possibly also 
calculate the integrated intensity for a background region and subtract 
it from the value of each band.

If you can't find a way to quantify the integrated intensity of the 
bands, then you could calculate the average or mean intensity. This will 
only work if the rectangles around your bands are identical, otherwise 
you are dividing the sum of pixel intensities for each band by a 
different denominator (number of pixels).

One thing to note: in all drawing programs, including Photoshop and 
ImageJ, white has a value of 255 and black has a value of 0. Therefore, 
you must quantify the negative of your image, where the background is 
black and the bands are white. Otherwise each pixel of background would 
add lots of signal (close to 255), while each pixel of band would only 
contribute a small amount (close to 0) to the total.

Hope this helps,
Irit


On 2/17/2011 5:47 PM, Joshua Silverstein wrote:
> I am also interested in knowing this so could you also send any replies to
> me as well? Thank you so much
>
> Josh Silverstein
>
> On Feb 17, 2011 8:34 PM, "Pragnya Das"<drpragnyadas from gmail.com>  wrote:
>
> I want to quantitate the bands from a Western Blot (X-ray film) and a PCR
> gel (Thermal paper). How can I do that. I have heard people using Photoshop
> to qunatitate. Is it scientifically accepted. If so, how do I proceed after
> scanning the films? I know a lot of labs use ImageJ software.
>
> Your suggestions will be greatly appreciated
>
> PD
> _______________________________________________
> Methods mailing list
> Methods from net.bio.net
> http://www.bio.net/biomail/listinfo/methods
> _______________________________________________
> Methods mailing list
> Methods from net.bio.net
> http://www.bio.net/biomail/listinfo/methods



More information about the Methods mailing list