The function computes for each cell the median CV and stores them
accordingly in the
colData of the
QFeatures object. The CVs in
each cell are computed from a group of features. The grouping is
defined by a variable in the
rowData. The function can be
applied to one or more assays, as long as the samples (column
names) are not duplicated. Also, the user can supply a minimal
number of observations required to compute a CV to avoid that CVs
computed on too few observations influence the distribution within
a cell. The quantification matrix can be optionally normalized
before computing the CVs. Multiple normalizations are possible.
medianCVperCell( object, i, groupBy, nobs = 5, na.rm = TRUE, colDataName = "MedianCV", norm = "none", ... )
Additional arguments that are passed to the normalization method.
A new column is added to the
colData of the object. The samples
(columns) that are not present in the selection
i will get
assigned an NA.
Specht, Harrison, Edward Emmott, Aleksandra A. Petelski, R. Gray Huffman, David H. Perlman, Marco Serra, Peter Kharchenko, Antonius Koller, and Nikolai Slavov. 2021. “Single-Cell Proteomic and Transcriptomic Analysis of Macrophage Heterogeneity Using SCoPE2.” Genome Biology 22 (1): 50.
data("scp1") scp1 <- filterFeatures(scp1, ~ !is.na(Proteins)) scp1 <- medianCVperCell(scp1, i = 1:3, groupBy = "Proteins", nobs = 5, na.rm = TRUE, colDataName = "MedianCV", norm = "div.median") #> Warning: The median CV could not be computed for one or more samples. You may want to try a smaller value for 'nobs'. ## Check results hist(scp1$MedianCV)