For many ELISA's your standard curve will appear nonlinear (kind of S-shaped), thus you must fit to the data using a nonlinear regression model.
I use GraphPad Prism software, they have online tutorials for interpolating from a standard curve. http://www.graphpad.com/guides/prism/6/curve-fitting/index.htm?reg_interpolating_from_a_standard_.htm
The steps are to enter data with your unknowns below the standards (just with no Y value), convert to a log format (X=Log(X)), do a nonlinear regression model (I use dose-response curves) making sure to check 'interpolate unknown values', then convert the values back into concentrations (X=10^X).
I doubt that you will be able to find such program. Any Elisa will give you a relative measurement, meaning that you will need to include known concentrations of the substance you are testing as standards, let's say 0g/l or blank, 0,25g/l, 05g/l, 0,75g/l and 1,0g/l. Then you can plot these known concentrations to the absorbance you measure, and with regression will be able to calculate your unknown concetration, as long as it's between 0,1g/l and 1,0g/l. These are just examples, you can pick your standards anywhere between the range of the test itself (check for minimum and maximum measuable c.
I would use Excel. Plot your results with the known concentration, apply regression and calcuate your unknown c. Here's a nice detailed page on how to do it. https://www.ncsu.edu/labwrite/res/gt/gt-reg-home.html
irrespective of Std concentration aavailability, OD values varies per assay. You have to make standard curve for each and every assay irrespective. Otherwise, the results will be erranious. Its mandatory to have standard references to be added for each assay.
For many ELISA's your standard curve will appear nonlinear (kind of S-shaped), thus you must fit to the data using a nonlinear regression model.
I use GraphPad Prism software, they have online tutorials for interpolating from a standard curve. http://www.graphpad.com/guides/prism/6/curve-fitting/index.htm?reg_interpolating_from_a_standard_.htm
The steps are to enter data with your unknowns below the standards (just with no Y value), convert to a log format (X=Log(X)), do a nonlinear regression model (I use dose-response curves) making sure to check 'interpolate unknown values', then convert the values back into concentrations (X=10^X).
Thanks Natalie, we've been using Excel but will give Prism a go.
Sorry - I meant to say "The steps are to enter data with your unknowns below the standards (just with no X value)"
I am using curve expert 1.4. it was giving very good results, after checking with different models. you try that one.
- Badges
- Science method
- Similar topics
- Medicine
- Immunology
- Immunology Techniques
- Immunoassay
- ELISA