ELISA data analysis
Posted 27 December 2009 - 01:32 AM
This is a really simple basic question for which most of you will have an answer but it will be conflicting as there are many ways to do and look at an analysis for a ELISA result.
If we do an ELISA of serum sample against a viral protein and try to see if there is a response to the viral protein, would you take it as positive if there is a 4 fold rise when compared to blanks which has no viral protein but every other reagent or will you consider 4 fold rise after you have minus ed the OD value from the negative which is the serum sample alone but no viral protein on the plate and than compare it to the plate negatives which is only reagents???
And if this is correct how would you plot your data, if it is looking at a population data? what would be the best way to interpret the data?
Posted 28 December 2009 - 05:25 AM
You can subtract out the blank well reading (no samples/serum just reagentes) from all samples tested.
Graphically, I would first test 20 healthy negatives and plot the distribution curve of the results. I would then plot the distribution curve of responses of those who seroconverted. These two normal distributions should not overlap except at the tail end where sero conversion starts.
You can also plot average response v. time during the seroconversion... ie days/weeks v. signal intensity/concentration.
If you are doing vaccination you may have cases in which there are non-responders.
Posted 28 December 2009 - 11:46 PM
We get pre- and post immunization sera and need to compare two immunogens, which is giving better seroconversion. Various 2-fold serial dilutions (starting 1:100) are used in the assay for both pre-immunization and post immunization sera and as a result we get respective OD values with satisfactory linearity. The assay includes a buffer blank also where we do not add serum but the diluent, all other reeagent are used in these wells.
From the values the data gives an indication that there is a seroconversion and apparently which immunogen is better. I would like to know the scientifically accepted way to calculate titer for both Pre and post immunization sera so that inference on comparison of two immunogens prepared by two different ways is justified.
I could find one reference where a titer was defined as reciprocal of highest dilution of serum showing OD more than Mean+2SD of a negative control. What does this Negative control mean, pre-serum or buffer blank, if pre-serum which dilution should be considered as we get high values from undiluted pre-immunization sera also, possibly due to natural exposure.
Looking for all possible answers and suggestions with thanks in advance.
Posted 29 December 2009 - 10:21 AM
You will have to titer all the normals you have and follow the course of the responders. On average how significant is the change 4 fold or greater? How many responded to the vaccine? one dose or more...igm to igg class switch.