Skip to content

Plot Boxplots by Subarray (Sample)

Source: R/boxplotSubarray.R
boxplotSubarray.Rd

Plots the distribution of all analytes, stratified by subarray, as a boxplot. These plots are intended to be used as a quality control visualization tool for SomaScan assay runs. In SomaScan (soma_adat) data format, the term "subarray" is analogous to sample, and typically indicates a row in the data.

Usage

boxplotSubarray(
  .data,
  color.by = NULL,
  labels = "SampleId",
  y.lim = NULL,
  do.log = TRUE,
  apts = NULL
)

Arguments

.data

A soma_data or data frame object, created from a SomaScan ADAT file, via a call to read_adat(). This object must contain the following columns: PlateId, SampleId, SampleType, SampleMatrix, Barcode2d, SlideId, Subarray, and HybControlNormScale.

color.by

Character. A column name to color the subarrays (samples) by. This is typically a sample processing or clinical data field in the ADAT such as SlideId.

labels

Character. The column name of .data used to label each box.

y.lim

Numeric. Length 2. The upper- and lower-quantiles of the total data used to determine the y-axis limits of the plot. If NULL, all points are shown.

do.log

Logical. Should the data be log10-transformed?

apts

Optional. A subset of analytes (as AptNames) to add as points on top of the subarray boxplot.

See also

geom_boxplot()

Other boxplots: boxplotBeeswarm(), boxplotGrouped()

Author

Stu Field

Examples

data <- SomaDataIO::example_data

# Randomly select a small subset of samples
s_rn <- withr::with_seed(101, sample(rownames(data), 30L))
data <- data[s_rn, ]
boxplotSubarray(data)


# Color by `SampleType` variable
boxplotSubarray(data, color.by = "SampleType")


# Find the feature names of the corresponding hyb controls
hybs <- SomaDataIO::getAnalyteInfo(data) |>
  dplyr::filter(grepl("^Hybridization", Type)) |>
  dplyr::pull(AptName)

# Pass hyb controls to `apts` arg; plots the points in boxes/whiskers
boxplotSubarray(data, color.by = "SampleType", apts = hybs)


# Zoom to (20, 80) quantiles
boxplotSubarray(data, color.by = "SampleType", y.lim = c(0.2, 0.8))


# Group by an additional variable
boxplotSubarray(data, color.by = "SampleType") +
  ggplot2::facet_wrap(~PlateId)