Example

Below is a fictive complex dataset named xplr.data and available in the R package. This dataset is a typic table obtain with metabolomic tools: we have 46 variables (métabolite amounts/peak areas) and 3 factors (Population, Temperature and Pesticide). For each of the 3 factors, we have at least 2 modalities/possibilities.

• Population = Africa; Finland or France
• Temperature: 25 or 36°C
• Pesticide: A; B or C

Thanks to xplR you can quickly calculate the means, standard deviationand errors of a set of several response variables in function of one or several experimental factors. For this user have just to give the name of response variable and factor they want to investigate. The output is a R list of numeric object.

• The function mean.list calculate the mean
• The function sd.list calculate the standard deviation
• The function se.list calculate the standard error
• The function xplr calculate all means, standard deviations and errors
• The function save.xplr save the xplr output in a csv file

For instance using our fictive dataset we measured these parameters for all metabolites among the three factors as you can below using the function save.xplr:

> save.xplr(xplr_data[,4:49],paste(xplr_data$Population,xplr_data$Pesticide, xplr_data$Temperature)) 

3.1/ Boxplots

You can create boxplot of your dataset to graphically explore your dataset and identify outlier and first variability trend among factors. xplR proposes 3 functions:

• xplr.boxplot to create boxplots of response variable among one or several experimental factors directly in the R window
• xplr.boxPDF to create a single PDF file containing boxplots of response variable among one or several experimental factors
• xplr.boxPDFs to create multiple PDF files containing each one the boxplot of one response variable among one or several experimental factors. The name of the file corresponds to the name give to each variable.

>xplr.boxplot(xplr_data[,10:13],paste(xplr_data$Population,xplr_data$Temperature),10, las=2, border="grey", col=c("lightblue",  "lightblue", "lightgreen", "lightgreen",  "lightpink", "lightpink"),ylab="mean amounts")

Here we investigated the dataset variation of 4 metabolites among two factors (Population and Temperature)

3.2/ Barplots

You can create bar plot to represent the mean and standard error of response variables from your dataset among factors. Likewise xplR proposes 3 functions:

• xplr.barplot to create bar plots of response variable among one or several experimental factors directly in the R window
• xplr.barPDF to create a single PDF file containing bar plots of response variable among one or several experimental factors
• xplr.barPDFs to create multiple PDF files containing each one the bar plot of one response variable among one or several experimental factors. The name of the file corresponds to the name give to each variable.

>xplr.barplot(xplr_data[,12:15],paste(xplr_data$Population,xplr_data$Temperature),12,las=2,border="black", col=c("lightblue","lightblue", "lightgreen", "lightgreen", "lightpink", "lightpink"),ylab = "Mean", offset=0)

Here we investigated the mean and standard error variations of 4 metabolites among two factors (Population and Temperature)

3.3/ Lineplots

You can create lineplot to represent the mean and standard error of response variables from your dataset among one factor (x axis label) and a group of several factors (line series). Lineplot are particularly intresting for kinetic dataset. Likewise xplR proposes 3 functions:

• xplr.lineplot to create lineplots of response variable among one factor for one or several group of factors directly in the R window
• xplr.linePDF to create a single PDF file containing line plots of response variable among one factor for one or several group of factors
• xplr.linePDFs to create multiple PDF files containing each one the line plot of one response variable among one factor for one or several group of factors. The name of the file corresponds to the name give to each variable.

>xplr.barplot(xplr_data[,4:15],paste(xplr_data$Population,xplr_data$Temperature),12,las=2,border="black", col=c("lightblue","lightblue", "lightgreen", "lightgreen", "lightpink", "lightpink"),ylab = "Mean", offset=0)

Here we investigated the mean and standard error variations of 4 metabolites from three different population (group=serie) among the type of pesticide they were exposed (factor)