# Interpretation of ELISA Data\Producing a Standard Curve - (Nov/29/2007 )

I'm a co-op student with my university, and am currently on my second practicum as a molecular biology undergrad. I've never done an ELISA before in an academic setting, let alone professionally, however am trying to develop one for my employer.

I was wondering if anyone could at all help me out with how I'm supposed to interpret the data. I've done other protein assays, BCA's, Bradford's etc. so I'm familiar with plotting a standard curve and using the curve to estimate the concentrations of other samples, however was wondering if it works in a similar manner for the ELISA. I know the curve is sigmoidal in shape, but is excel supposed to be able to accurately find the equation for the curve by regression or is it more or a visual interpretation/estimation, when using the curve to determine concentrations of "unknowns"/tested samples?

Any advice or tips would be greatly appreciated.

-Jordan Cran-

The standard curve is your best friend in an ELISA! Because you're just looking at the same protein, all you have have to do is read off the OD of your test sample. Couldn't be easier, or nicer!
As to using Excel, you don't usually need to define the resultant curve for data interpretation.

-swanny-

Hello!
Currently I am using an IL-8 Elisa kit from R&D. I just saw in the manual that they plot the logs of the O.D. and the concentrations againts each other. Does this give me any advantage? I am a bit confused as I always did a normal standard curve with the concentration on the X axis and the O.D. on the y axis. Was I wrong?

-Joohn-

For the normal X-Y plot, do you advise to :
- draw the best fitting line through the point (0;0) of the plot or not?
- do you subtract the O.D. of the 0 pg/mL from all the other O.D.?

-Joohn-

1. Log plots turn the sigmoidal curve into a (hopefully) lineal one, which makes reading results much easier. Concentration is on the X, and OD on the Y axis, because the OD is dependent on the concentration, and the dependent variable is always on the Y.

2. When plotting the curve, I presume you did a blank. The value you get is the value you should use. It's a bit academic, because if your results are down that part of the curve, you have so little that you should consider changing the protocol a bit (try extending the final colour development step, until your sample ODs become meaningful. You might also need to adjust the range of your standard curve to get useful data for plotting.

-swanny-

QUOTE (swanny @ Feb 29 2008, 08:12 AM)
1. Log plots turn the sigmoidal curve into a (hopefully) lineal one, which makes reading results much easier. Concentration is on the X, and OD on the Y axis, because the OD is dependent on the concentration, and the dependent variable is always on the Y.

2. When plotting the curve, I presume you did a blank. The value you get is the value you should use. It's a bit academic, because if your results are down that part of the curve, you have so little that you should consider changing the protocol a bit (try extending the final colour development step, until your sample ODs become meaningful. You might also need to adjust the range of your standard curve to get useful data for plotting.

I agree that log plots is nice for that curve, then you have OD of your sample, put it in Y value and select one X value, then see whether it is in the line or not. If not, select a better value of X, until you can choose the best one.

-Malaria-