# Concentration calculation of Factor VIII for Elisa

I have the following problem. I'm trying to get the concentration calculation for an Excel sheet that will calculate values from an Elisa plate.

The basis is a standard curve with the following values:

Std: PreDil factor 1: 40 PreDil factor 2: 55 ---> Total Predilution = 2200 Concentration: 2123 µg/ml

Dilution 1:20 1:40 1:80

Std 1 1,3910 0,6360 0,2850

Std 2 1,4020 0,6360 0,3140

Std 3 1,3500 0,6250 0,2880

Std 4 1,4470 0,6810 0,2900

The sample values are the following:

Predilution: 15

Dilution 1:20 1:40 1:80

Sample 0,4060 0,1940 0,0760

According to the program that has been in use earlier, the concentration shall become 4,786 µg/ml. How do I calculate this in the best way? Unfortunately, I can't seem to get the right answer.

To add more context to the question. I work on a project to move an old program that calculated the concentrations to a new excel sheet. Unfortunately, I do not perform these analyses myself as I work on another part of the lab.

However, the question is based on how to get the concentration value of the undiluted sample. The standard used have an undiluted concentration of 2123 µg/ml. It is pre-diluted in two steps, first 1/40 and then 1/50 giving a total dilution of 1/2200.

The Standard is then analysed for three different doses 1/20, 1/40 and 1/80. Two times at the beginning of the run, and two times in the end.

The Sample is pre-diluted 1/15 to then be analysed for the same three doses as the standard.

What I have done now is that I used the Parallel Line model to try to calculate the concentration. Unfortunately, I don´t get the same answer as the earlier calculation program (Which is 4,786), and I have been able to figure out is that the problem is that I get a slight shift in the relative potency. Unfortunately, I can´t find where the problem lies.

Hopefully, this clears it up a small bit.

• This just doesn't make sense. You say that the undiluted sample has a higher concentration than the undiluted standard. The assay values are of the same order of magnitude for standards and sample. But the standards were prediluted x2200 while the sample was only prediluted x15. If the sample is more concentrated and is prediluted less then it should give much higher readings. Or I don't understand your experiment. – Alan Boyd Jul 9 '17 at 13:37
• Hi sorry for the late reply. The undiluted standard got a concentration of 2123µg/ml while the sample got a concentration about 4,8 µg/ml. – Martin Jul 11 '17 at 5:20

## 1 Answer

My answer: 4.79 µg/ml

(All values below are µg/ml)

Calibration data

Determine the mean of the four absorbance values at each dilution.

The calibration curve is then a plot of concentration (x-axis) against mean absorbance. Note that the actual concentration in the assay is (2123/2200)/dilution where dilution is 20, 40 or 80. If you chart this you will see that the data fit well to a straight line.

Use the Excel SLOPE and INTERCEPT functions to derive the parameters of the trendline. (I get SLOPE = 30.59; INTERCEPT = -0.8225)

The equation for converting an observed absorbance value to a concentration is:

concentration = (absorbance - INTERCEPT)/SLOPE

Sample data

For the sample measurements the only one that is valid is that for the 20x dilution because the other absorbances lie outside the range of the calibration data (see note below).

Using 0.406 as the absorbance value in the equation above gives a value of 0.01596 for the concentration in the assay well. This sample was diluted 20-fold for the assay, and the source for that dilution was derived from a 15-fold ‘pre-dilution’ so the original concentration of the sample is 300x the assay value which is 4.788 µg/ml.

Note

I don't know the nature of the discrepancy between your calculations and the target value, but if you try to incorporate the 40x and 80x dilutions for the sample you will get higher values for the sample concentration. Presumably the assay becomes non-linear at low concentrations.

• Hi Alan, Thank you´so much for the response. It really helped a lot and I have been able to work out how to perform this when only one of the samples are in range. Now I´m stuck instead on how to incorporate the values when all of them are in range. Example: Predilution for standard = 1, conc = 1,04 Avg ystd = 1,1585 0,611 0,3205 Y sample = 1,16 0,687 0,345 with a dilution of 30. I try to put them on a common slope, get a value for the relative potency and then calculate the concentration. The problem is that I get a value of 31,88 when the "correct" result is 33,50. – Martin Jul 11 '17 at 11:07
• @Martin Not sure that I fully understand the numbers you mention, but if you have more than one assay of the sample that is in the range of the standard curve then you can simply use all valid readings to derive separate measurements of the [sample], remembering to vary the dilution factor as appropriate (e.g. x20 above but if the 40x dilution had also been valid then use the same formula but multiply by 600 (15x40) at the last step). You can then take the average value of all of your estimates of [sample]. (Apologies for using 'mean' and 'average' interchangeably here.) – Alan Boyd Jul 11 '17 at 11:09
• Thank you so much, Alan! Got confused with the common slope that is mentioned in my papers, this helped me out fantastically! – Martin Jul 11 '17 at 11:33