Perform Weighted Linear Regression for Calibration Curve.

Welcome to CCWeights

Calibration curves are used to understand the instrumental response to an analyte and predict the concentration in an unknown sample. A well-established and interpreted calibration curve is essential for any analytical methodology.

CCweights is a web-based tool (also an R package), which provides automated and efficient data analysis workflow to evaluate and select the best weighting factor for linear calibration curve and quantify targeted analytes accordingly.

CCWeights can be applied for any analytical assays, in which linear calibration curves are used for quantification. For instance, Ultraviolet-visible spectroscopy- and liquid/gas chromatography mass spectrometry-based quantitative studies.

Installation

1. stable version

install.packages("CCWeights")

2. development version

require(devtools)
install_github("YonghuiDong/CCWeights")

Note: you need to install R package bs4Dash version <= 0.50 in order to run CCWeights shiny app locally. Please refer to section Known Issues for more details.

Workflow

Below is an overview of CCWeights workflow:

Figure 1. Schematic workflow of CCWeights

Data Preparation:

• Data must contain at least two columns, one is Concentration, and another one is Response. In case you have more than one compound in your sample, you need to have a third column called Compound. If internal standards are used (e.g., stable isotope labeled internal standards), please include another column named IS, with the corresponding response value filled in the cell.

• Use unknown for compounds with Concentrations to be determined in samples.

• Data with known Concentration will be regarded as calibration standards for calibration curve construction and evaluation.

• Data with unknown concentration will be regarded as unknown samples, and the Concentration will be predicted.

• You can refer to the example data in the Upload Data tab or Figure 2 for the data format.

• You can also download the templates here to prepare your data accordingly.

Figure 2. Data format requirement

• Once data file is successfully uploaded, you can perform data analysis following the steps shown in Figure 1.

Known Issues

Due to substantial breaking changes in the API of R package bs4Dash, CCWeights, built with bs4Dash (version <= 0.5.0), is not compatible with v2.0.0. Therefore, users need to install the old version in order to run the shiny App of CCWeights locally.

There are two simple ways to install the old version of bs4Dash:

1. Using devtools to install the older package version

require(devtools)
install_version("bs4Dash", version = "0.5.0", repos = "http://cran.us.r-project.org")

2. Installing an older package from source

You can download the old version (e.g., v0.5.0) here, and install it manually.

Citation

If you find CCWeights useful, please consider citing our publication :smiley::

• CCWeights: An R package and web application for automated evaluation and selection of weighting factors for accurate quantification using linear calibration curve.