Band Intensity - Linear or Exponential? PCR Products - (Mar/01/2008 )
Below is a gel I ran for some PCR products.
The last two on the right-top and the last two before the ladder on the bottom are different dilutions of input for a ChIP experiment. 1:5, 1:25, 1:125, and 1:625
Strangely when I use ImageJ, get a histogram, take the area under the curve for the 'bands' (subtracting baseline), I get a very strong curve fit for the samples - but it is exponential, I thought it should be linear (though I don't know why I think that)?
I could swear, in the past, I have adjusted the amount of input DNA I used in certain PCRs by calculating band intensity of Actin PCR divided input DNA, then adjusted all the input DNA amount to rerun the PCR, and got equal intensities across the board. But this would be linear changes, right? (maybe because the intensities were pretty close to begin with, it worked?)
What do people think?
Shawn McClelland
www.e-lab-book.com
Or shouldn't it be sigmoidal, but the linear area is where it's reliable? I mean my curves fit R2 of .998 for exponential.
You have to ensure with your software before you get the picture that you are not overexposing any area. Most image capturing programs nowadays have a control that enable people to see where there is an overexposure . In this case you reduce the exposure time to levels that none of your bands of interest have been overexposed. Judging by your image the two bands at the bottom left corner, the seventh band in this lane and the two on the top right corner have been overexposed.
?? You didn't really answer my question, but thanks for the input.
In my opinion the bands aren't overexposed. The analysis curve fits absurdly well (not linearly though - which was the qeustion I am an asking). The only bands I am looking at are the two on the top right and the two on the bottom right - before the ladder.
?? You didn't really answer my question, but thanks for the input.
In my opinion the bands aren't overexposed. The analysis curve fits absurdly well (not linearly though - which was the qeustion I am an asking). The only bands I am looking at are the two on the top right and the two on the bottom right - before the ladder.
In my view, the bands are overexposed. The reason you are getting this curve is probably because the software attributes a bigger area to the most overexposed bands as there is a diffusive effect of the extra light in the picture. It is interesting, however, if under this conditions you are getting an exponental relationship.
regards
odiporos