You can compare your samples and controls using absorbance (see if they are different - non-parametric) -  however, as Mahmoud point out, you do not know the concentration in your samples without the standard (and standard-curve tells you that assay works properly). Also, you need to use the standard if you will compare results obtained on different days as absorbance may vary from day to day.

If you do not use the standards how you will be able to determine your results?

I'm not sure as it's the first time i am doing something like that. Can i determine my results from the absorbances? Can i make a comparison between my samples (with the different concentrations of the substance i examine) and my controls? Is that possible?

The standard curve is the curve between the known concentrations of the substance and absorbance. So You can not know the concentration of your testing substance without standard curve

Please dilute your sample if you expect that concentration is high (outside standard Curve)

You can compare your samples and controls using absorbance (see if they are different - non-parametric) -  however, as Mahmoud point out, you do not know the concentration in your samples without the standard (and standard-curve tells you that assay works properly). Also, you need to use the standard if you will compare results obtained on different days as absorbance may vary from day to day.

If you are not sure about the dilution factor of the samples, you could run the assay without the standards (incl blank). Dilute the samples to get an OD value in between 0.4 and 1. Assay along with standards to get the actual concentration. 

OD values will be different if you change the incubation time/ temperature/ expose to light. So it is always advisable to run standard along with the sample to get the actual concentration. I hope this answers helps.

The comparison you talk about (OD values of controls vs OD values of samples) would give you a qualitative result: which samples has higher or lower absorbances. It is of little use without the standard curve, as OD values cannot be analyzed properly by a statistic package, they need to be converted to (in this case) concentration values (pg/ml) to do so. If I were you, I would not waste those wells.  

When working with cell culture, it is recommended to dilute the sample supernatant prior to performing the ELISA (kits provide dilution buffer to do so), specially with TNF-a. Check for similar studies measuring TNF-a in order to apply a proper dilution factor so your results come within the standard curve.

Good luck       

Yo need to use the curve, otherwise you wont be able to obtain a strict comparison of your samples beyond gross differences.

In short, if you simply want to see if you can get a reaction with your samples and/or how strong this reaction is, you will not need the curve.

for your actual experiments: If you want to measure and quantify the concentration in your samples you need the curve to calculate it. As mentioned before, you will need a curve on every assay you run because conditions between assays may vary.

I do not understand, why is here everybody so opposed to running the samples without the standards, if the purpose is to only preliminary compare samples (and see how it works). Why are you so opposed to comparing the OD?

You will convert all ODs in the same way by one formula anyway, so the differences will be the same for the statistical analysis or for seeing the differences between samples.

I agree with Dres Damgaard and Leen Jansen, if Ioanna wants to see preliminary results, she may well proceed without the standard curve. However, for any comparison between plates (and any actual publication), the standard curve will be necessary.

Hello everyone,

I have a related question. So I have an ELISA kit and the standard ran out, but I need to run an urgent assay. I have ran standard curves with this kit many times.

Is it ok to run the assay and use one of the old standard curves to determine my values?

Thanks.

In general the answer is No. If you want to quantitate the levels of your analyte you need the standard curve in the same assy.

But, if you have done it many times and the results are really reproducible (same Abs values for the standards every time) you could take the risk, even knowing that this is not the ideal way to do it because perhaps the new ELISA is going to be lower or higher than the previous ones. Nobody can predict that. On the other hand, if the curve had variable values in every assay, as it often happens, you cannot even think in doing that.

aha. That makes sense. Thank you for your answer.

My question is if it is possible to analyse ELISA results similarly to what we do by WB. I have two ELISA kits for 2 different proteins. One of them is the internal control, IC (like actin or tubulin in a WB) and the other is my protein of interest, PI, susceptible to changes. The same sample will be run then using the 2 kits in paralel. I am interested in the ratio between the absorbances for both proteins in the same sample, and then compare the control-sample values to the different experimental conditions. I understand that the quantitave absolute value can not be obtained unless you use a standard curve, however I want to know the relative abundance of my PI/IC among the different conditions to be analysed. Most of publications just show changes in the total amount of a single protein assuming equal loading for all conditions. This is not a suitable case for us and we need to relativize to a loading control. Will absolute values after a STD curve give me any extra info than the ratio of absorbances? Can I just not measure relative absorbances and give 1 to control and compare to the other experimental conditions? wouldn't the "proportionallity" of changes in total amount of protein or the ratio of absorbance be the same?

Thanks!

Hi Ana, in ELISA the calibration curve generally is not linear so, if you stick to the absorbance only, you will end up into troubles.

The standard curve has to be linear. That is why values below or above are not good, they scape from the linearity of the reaction right?

No, the standard curve is not always linear, you can use a logarithmic regression (or some other) as well to calculate concentrations by ELISA. And without running a standard curve, it will be difficult to know whether your high value samples are similar or just in the saturation range of the assay, where differences are not detected anymore.

If you dilute the samples enough to avoid saturation you could use the Abs ratio to compare samples within the same plate. But aAs ratio cannot be used to compare samples from different ELISAs. In this case, you would need the standard curves in each assay.

Thank you all for your help!!!