Dear all,
I am running multiple ELISA plates from invitrogen (same kit and same batch number). In each plate I run 5 standards plus one blank and all plasma samples in duplicate.
I tried to exam the precision of immunoassays by examining reproducibility of reported results through examining variation between wells within a single run of a plate (intra-assay precision) or between runs or plate-to-plate (inter-assay precision).
I had quite good intra-assay precision, but when run the same samples in different days I saw different concentration.
I investigated and I see the CV% of standards some days is quite low and some days quite high.
I wonder to know how I should normalised my samples? I google and I found out some answers as below:
1- Use one standard curve for all plates (I will choose the best standard curve with lowest CV% and I will use for all 25 plates?)
2- By dividing the OD of the test sample by the OD of the blank, results can be expressed as a fold change. And then what should I do?
I appreciate your help and comments.
Hi, no I did not. What I am running is sandwich ELISA which coated with antibody in the bottom of each well. The total protein concentration do not interfere with the interest biomarker (protein) concentration.
I wouldn't recommend using standard curves from a different plate for an ELISA. If there are high CV values either the homogeneity of the plate is the cause or some errors in the pipeting and/or washing steps.
Quite often the washing step is performed very differently in different labs (very quickly for a few secondes manually with variable incubation times at each step or slowly with automated washer, but different residual buffer volumes) and shows most variations with high background etc.
An additional source of heterogeneity is the coating of the plate where you can get high variations, mainly with more sensitive antibodies where the drying process of the plates can lead to random wells with high signal differences or different drying on the edges of a plate where inner wells show higher homogeneity than at the edges.
You can't change the manufacturing issues but could try to have a more standardized way of pipeting. Plus, using triplicates instead of duplicates for the standards and samples could help masking outliers and still get decent CVs for standards.
- Badges
- Science topic