I have been testing 2 different plasma dilutions to see which would be better to use moving forward in a competitive ELISA assay. The issue I am having is that there is a 2.5x-3.75x fold-difference between the calculated final concentration of each sample after I correct for the dilution factor. I expect the concentrations to be similar. The 1:5 samples are also showing higher concentrations (lower OD values) than the 1:2 diluted samples.
I have ran 4 samples in duplicate, each with a 1:2 and 1:5 dilution. The data was analyzed using Calbiotech's Analyzing Data for a Competitive ELISA video. R^2 = 0.9999 with second order polynomial trend line.
Is there any explanation to this minus possible pipette error?
How do I know which dilution to move forward with if the OD values for both dilutions are in the middle of the standard curve but the final concentrations are so different?
Any input would be much appreciated.
One possible explanation is that the range of correct conc. determination of an ELISA is between an exponential/better a linear area and the plateau phase (which may be reached a too high protein conc. in some of your samples/lower dilutions). Whereas, results from higher diluted samples lay in a more adequate range (near to a linear phase). This may lead to different outcomes regarding the calculated "total" concentration (dependig on the standards). In general I would propose to check the manufacurers protocoll for those information on most useful as well as max. protein concentrations, the ELISA was designed for. Additionally, I would propose to contact customer service in this context.
Oh... it is too complicate question for a fast explanation.
First, you didn't tell what are you detecting (Antigen or antibody). Next You did not tell what was difference between results? Maybe it was near differences between duplicates.
I can take advice you to pay attention on two things:
1. The 2-nd polynomial interpolation isn't the best way for calculation of ELISA's results. If you want get more accurate results then you have to use the 4- or 5-parametric logistic model.
2. Analits in the standard and the sample can have various nature. (whole virions and soluble antigen, one- and two chain antibodies etc). In this case curves of dilution of these samples will be discribed by different functions.
The best way will be using serial dilutions of your samples and calculating of the concentration by the parallel lines method or the 4- or 5-param logistic regression.
you can see also https://www.bio-rad-antibodies.com/elisa-results-quantitative-semi-qualitative.html
https://www.oxfordbiomed.com/sites/default/files/2017-02/How%20to%20Obtain%20Reproducible%20Quantitative%20ELISA%20results.pdf
I think you used one standard curve for calculations of both the dilutions. If any one dilution is out of the standard curve range then this type of problem will arise.
- Badges
- Science topic
- Similar topics
- Microbiology
- Assays