aggr_rep: Aggregate "Sample__Condition" Replicates in Data Matrix

In spatialHeatmap: spatialHeatmap

Description

This function aggregates "sample__condition" (see data argument) replicates by mean or median. The input data is either a data.frame or SummarizedExperiment.

Usage

aggr_rep(data, sam.factor, con.factor, aggr = "mean")

Arguments

data	An object of data.frame or SummarizedExperiment. In either case, the columns and rows should be sample/conditions and assayed items (e.g. genes, proteins, metabolites) respectively. If data.frame, the column names should follow the naming scheme "sample__condition". The "sample" is a general term and stands for cells, tissues, organs, etc., where the values are measured. The "condition" is also a general term and refers to experiment treatments applied to "sample" such as drug dosage, temperature, time points, etc. If certain samples are not expected to be colored in "spatial heatmaps" (see spatial_hm), they are not required to follow this naming scheme. In the downstream interactive network (see network), if users want to see node annotation by mousing over a node, a column of row item annotation could be optionally appended to the last column. In the case of SummarizedExperiment, the assays slot stores the data matrix. Similarly, the rowData slot could optionally store a data frame of row item anntation, which is only relevant to the interactive network. The colData slot usually contains a data frame with one column of sample replicates and one column of condition replicates. It is crucial that replicate names of the same sample or condition must be identical. E.g. If sampleA has 3 replicates, "sampleA", "sampleA", "sampleA" is expected while "sampleA1", "sampleA2", "sampleA3" is regarded as 3 different samples. If original column names in the assay slot already follow the "sample__condition" scheme, then the colData slot is not required at all. In the function spatial_hm, this argument can also be a numeric vector. In this vector, every value should be named, and values expected to color the "spatial heatmaps" should follow the naming scheme "sample__condition". In certain cases, there is no condition associated with data. Then in the naming scheme of data frame or vector, the "__condition" part could be discarded. In SummarizedExperiment, the "condition" column could be discarded in colData slot. Note, regardless of data class the double underscore is a special string that is reserved for specific purposes in "spatialHeatmap", and thus should be avoided for naming feature/samples and conditions.
sam.factor	The column name corresponding to samples in the colData of SummarizedExperiment. If the original column names in the assay slot already follows the scheme "sample__condition", then the colData slot is not required and accordingly this argument could be NULL.
con.factor	The column name corresponding to conditions in the colData of SummarizedExperiment. Could be NULL if column names of in the assay slot already follows the scheme "sample__condition", or no condition is associated with the data.
aggr	Aggregate "sample__condition" replicates by "mean" or "median". The default is "mean". If the data argument is a SummarizedExperiment, the "sample__condition" replicates are internally formed by connecting samples and conditions with "__" in colData slot, and are subsequently replace the original column names in assay slot. If no condition specified to con.factor, the data are aggregated by sample replicates. If "none", no aggregation is applied.

Value

The returned value is the same class with the input data, a data.frame or SummarizedExperiment. In either case, the column names of the data matrix follows the "sample__condition" scheme.

Author(s)

Jianhai Zhang [e-mail]; [e-mail]
Dr. Thomas Girke [e-mail]

References

SummarizedExperiment: SummarizedExperiment container. R package version 1.10.1
R Core Team (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/
Keays, Maria. 2019. ExpressionAtlas: Download Datasets from EMBL-EBI Expression Atlas
Love, Michael I., Wolfgang Huber, and Simon Anders. 2014. "Moderated Estimation of Fold Change and Dispersion for RNA-Seq Data with DESeq2." Genome Biology 15 (12): 550. doi:10.1186/s13059-014-0550-8
McCarthy, Davis J., Chen, Yunshun, Smyth, and Gordon K. 2012. "Differential Expression Analysis of Multifactor RNA-Seq Experiments with Respect to Biological Variation." Nucleic Acids Research 40 (10): 4288–97
Cardoso-Moreira, Margarida, Jean Halbert, Delphine Valloton, Britta Velten, Chunyan Chen, Yi Shao, Angélica Liechti, et al. 2019. “Gene Expression Across Mammalian Organ Development.” Nature 571 (7766): 505–9

Examples

## In the following examples, the 2 toy data come from an RNA-seq analysis on developments of 7
## chicken organs under 9 time points (Cardoso-Moreira et al. 2019). For conveninece, they are
## included in this package. The complete raw count data are downloaded using the R package 
## ExpressionAtlas (Keays 2019) with the accession number "E-MTAB-6769". Toy data1 is used as a
## "data frame" input to exemplify data with simple samples/conditions, while toy data2 as 
## "SummarizedExperiment" to illustrate data involving complex samples/conditions.   

## Set up toy data.

# Access toy data1.
cnt.chk.simple <- system.file('extdata/shinyApp/example/count_chicken_simple.txt', 
package='spatialHeatmap')
df.chk <- read.table(cnt.chk.simple, header=TRUE, row.names=1, sep='\t', check.names=FALSE)
# Columns follow the namig scheme "sample__condition", where "sample" and "condition" stands
# for organs and time points respectively.
df.chk[1:3, ]

# A column of gene annotation can be appended to the data frame, but is not required.  
ann <- paste0('ann', seq_len(nrow(df.chk))); ann[1:3]
df.chk <- cbind(df.chk, ann=ann)
df.chk[1:3, ]

# Access toy data2. 
cnt.chk <- system.file('extdata/shinyApp/example/count_chicken.txt', package='spatialHeatmap')
count.chk <- read.table(cnt.chk, header=TRUE, row.names=1, sep='\t')
count.chk[1:3, 1:5]

# A targets file describing samples and conditions is required for toy data2. It should be made
# based on the experiment design, which is accessible through the accession number "E-MTAB-6769"
# in the R package ExpressionAtlas. An example targets file is included in this package and 
# accessed below. 
# Access the example targets file. 
tar.chk <- system.file('extdata/shinyApp/example/target_chicken.txt', package='spatialHeatmap')
target.chk <- read.table(tar.chk, header=TRUE, row.names=1, sep='\t')
# Every column in toy data2 corresponds with a row in targets file. 
target.chk[1:5, ]
# Store toy data2 in "SummarizedExperiment".
library(SummarizedExperiment)
se.chk <- SummarizedExperiment(assay=count.chk, colData=target.chk)
# The "rowData" slot can store a data frame of gene annotation, but not required.
rowData(se.chk) <- DataFrame(ann=ann)

# Aggregate "sample_condition" replicates in toy data1.
df.aggr.chk <- aggr_rep(data=df.chk, aggr='mean')
df.aggr.chk[1:3, ]

# Aggregate "sample_condition" replicates in toy data2, where "sample" is "organism_part" and
# "condition" is "age". 
se.aggr.chk <- aggr_rep(data=se.chk, sam.factor='organism_part', con.factor='age', aggr='mean')
assay(se.aggr.chk)[1:3, 1:3]

spatialHeatmap documentation built on Nov. 8, 2020, 5:46 p.m.