Monthly Archives: June 2014

What a user should do if they have two standard curves on one plate?

We understand that most experiments may use only half or few wells of the plate.  Hence two or sometime three readouts with their respective standard curves can be designed into one single ELISA plate. Currently, allows users to insert only one standard curve per plate. In order to analyze layouts with two or more standard curves (Figure A) with ELISAAnalysis, the user must split the raw data manually into partial data sets and analyze the OD readings separately for each partial data set (Figure B).

Figure A:  Raw data in Excel

Figure A: Raw data in Excel

Figure B: Split data into two plate layouts in Excel

Figure B: Split data into two plate layouts in Excel

What does the option “starting reference number” mean in Step 2 when creating a layout?

The “starting reference number” relates to the next sample number the investigator is going to mark on the plate layout. It can either be the standard or the unknowns.  For example, in the plate layout if you have already marked wells as Unknowns 1-3, then the “starting reference number” should be set to 4 for the next Unknown sample (this should automatically happen). The same concept is followed for the standards. Soon we will be offering more powerful labelling in our plate layouts to specify detailed descriptions rather than just number references.

What should I do when I have controls/negatives for two different conditions on the same plate?

This experiment involves comparing pairs of treatment and control replicates on one plate. At present ELISAAnalysis software does not have the option of associating specific control/negative wells with specific treatment wells. Any wells marked as negatives are used to calculate the background cut-off for all unknowns on the plate.

There are two options for working around this limitation:

  1. Split the plate into two smaller data sets and analyse these data sets separately with
  2. Specify the controls and treatment wells as “Unknowns” in the plate layout. ELISAAnalysis will calculate the concentrations for both and the investigator can then manually compare the controls vs. treatment group in the results.

Which wells should be considered “blank” or “negative” while creating your layout?

ELISAAnalysis uses “Negatives” for calculating the Background Cut-off Threshold.  So if you feel that your control could be a factor for potential noise in the experiment, then you should specify your controls as “Negatives” in the plate layout. Please note that “Blanks” are completely ignored for calculations in our software.  A knowledge base article on background cut-off could be helpful for you before you proceed to calculation step and can be found here:

What to do if you need a standard curve without know concentrations of standard?

On unique occasions, the absolute value for standard for a particular ELISA is not available for generation of a standard curve. For example there is no purified native CD36 available as an absolute standard for CD36 human ELISA. Frequently, the investigator in these cases will prepare dilutions of a calibrator sample to help in estimation of the test analyte. The calibrator is usually a pooled sample prepared from combining all the samples to be tested in an experiment with unknown quantities of the analyte. The calibrator should be included on each plate and as it aids in comparing the results between different plates and also in quantification of the relative amounts of analytes. As an example, Figure A. shows the raw data where the standard is unavailable and the calibrator is used for the analysis by pooling serum samples from 4 human subjects.

Figure A: Raw Data in Excel

Figure A: Raw Data in Excel

The analysis of data in this format is currently not supported by ELISAAnalysis since the user can only insert concentration values and not dilutions. A simple workaround solution for handling these types of data is suggested below:

1. The dilutions of the calibrator in Figure A can be multiplied by a factor to get whole number values. Figure B depicts the concentration values obtained upon multiplication of calibrator by an appropriate factor (i.e. 819200). Then the raw data will look like Figure C.


Figure B: Conversion of calibrator dilutions to absolute concentrations in Excel.

Figure B: Conversion of calibrator dilutions to absolute concentrations in Excel.


Figure C: Plate Layout of the converted data in Excel

Figure C: Plate Layout of the converted data in Excel

2. Subsequently, the OD values obtained for the calibrator are inserted in the marked “standard” wells & test samples in the “unknown” wells in a custom plate layout or one of the popular layout of ELISAAnalysis (Figure D). After this 4PL analysis should be performed (Figure E).

Figure D:  Plate Layout of the converted data in ELISAAnalysis

Figure D: Plate Layout of the converted data in ELISAAnalysis

Figure E. Results from 4PL analysis of converted data where the concentration values of unknown (Test) has been calculated relative to the calibrator.

Figure E. Results from 4PL analysis of converted data where the concentration values of unknown (Test) has been calculated relative to the calibrator.

3. To get back to the dilutions of the calibrator we can divide the predicted values by the factor (in this case 819200). However this type of analysis will provide fold change expression, more like a qualitative measure. This is because without defining the relationship between protein concentration and response (which is what a standard curve does) any measurements will simply be relative.


How to Tighten ELISA Standard Curve Confidence Intervals

Now that provides 95% confidence intervals, many clients have asked how they can tighten the confidence intervals for their standard curves.  Here are some comments and suggestions:

  • You will usually see the confidence intervals widening for higher concentrations.  This is normal when the percentage variability is roughly constant for all concentrations.  For example, if the confidence interval is plus or minus 10% then for higher concentrations this will translate into a larger confidence interval.
  • For a given level of variability, to tighten the confidence interval you should increase the number of replicates for your standard curve points.  While duplicates for standard curve points are common, from a statistical point of view this is a very low level of replication and as a result this leads to low confidence.
  • For a given number of replicates, to tighten the confidence interval you should explore strategies for reducing the variability between replicates.  ELISA involves a large number of steps and the challenge is to implement experimental procedures to reduce variability at each step.
  • The 4PL curve has asymptotes at both the positive and negative extremes.  As the curve moves closer to the asymptote, the confidence intervals will widen.  For a given concentration, the tightest confidence intervals for a 4PL curve will be near the inflection point of the curve, where it looks roughly like a linear curve at 45%.  If your unknown values are near the asymptotes then you could consider diluting the samples to move them towards the centre of the curve or consider a more sensitive kit.

ELISA Standard Curve Confidence Intervals Log Scale

ELISA Standard Curve Confidence Intervals and our associated companies are really passionate about high quality science.  So to support our clients we decided to include 95% confidence intervals around the linear and 4PL standard curves. The region between the lower and upper confidence bound on the concentration axis is an estimate of the 95% confidence interval.  We hope that you find them useful!

ELISA Standard Curve Confidence Intervals

Here are some further comments and features relating to confidence intervals:

  • We understand that some clients might not want the confidence intervals on their charts.  To remove them, simply click on the legend titles for each curve.
  • The confidence intervals only show where the values on the curve are not negative.  So if only one curve shows on your chart this likely to be the reason.
  • You might notice that the “Upper Confidence Bound” appears to be lower than the “Lower Confidence Bound”.  This seems a little confusing until you notice that the confidence bounds are for the predicted concentration on the horizontal axis.
  • The confidence intervals have been calculated using standard tools in the drc package of the R statistical software.

To read about how to tighten your confidence intervals, please refer to this article: How to Tighten ELISA Standard Curve Confidence Intervals

If you have any questions please feel free to contact us!

Raw ELISA Data Input, ELISA OD Data Input

To input data into, simply copy the raw OD data from a spreadsheet or text file and paste it into the input box. will automatically recognise the data and format it into a 8×12 matrix with the same labelling as a standard 96 well plate.

ELISA Raw Data Input Box

ELISA Raw Data Input Box

Here are some further details on what you can do:

  • To edit a well simply double click it
  • If you paste in less than 96 wells then the remaining wells will be treated as blanks
  • The input box will recognise raw data seperated by commas, tabs or spaces with each row on a new line.

If you run into problems importing your data then please contact us as we are keen to ensure all standard formats work on our platform.