nCal - Some Examples

The function ncal carries out nonlinear calibration, which entails fitting concentration-response curves to sets of samples of known concentrations, also known as standard samples, and using the fitted curves to obtain point estimates and confidence intervals for the analyte concentrations in the samples of interest, also known as the unknown samples. Take Luminex (Defawe et al., 2012) as an example. The assay is conducted on 96-well or 384-well plates. Each plate usually has a set of wells containing standard samples and a set of wells containing unknown samples. Within each well, multiple analytes can be assayed simultaneously. In the following, we will use plate and assay exchangeably. ncal has 2 required arguments, formula and data, and 24 optional arguments. data is a data frame object, where each row is a sample, either standard or unknown. The formula object specifies the names of the response and concentration columns. Besides these two columns, data is also expected to have two columns that help identify standard curves: analyte and assay id. If data contains both standard and unknown samples, two more columns are also required: well role and sample id. data may also have additional columns. For example, it is sometimes convenient to include dilution factor for unknown samples. We now simulate a dataset with one assay, one analyte and four unknown samples with varying dilutions.