contrast_each_group_to_the_rest_for_norm_ma_with_limma: contrast_each_group_to_the_rest_for_norm_ma_with_limma

In MonashBioinformaticsPlatform/celaref: Single-cell RNAseq cell cluster labelling by reference

Description

This function loads and processes microarray data (from purified cell populations) that can be used as a reference.

Usage

  contrast_each_group_to_the_rest_for_norm_ma_with_limma(norm_expression_table,
  sample_sheet_table, dataset_name, sample_name, group_name = "group",
  groups2test = NA, extra_factor_name = NA, pval_threshold = 0.01)

Arguments

norm_expression_table	A logged, normalised expression table. Any filtering (removal of low-expression probes/genes)
sample_sheet_table	Tab-separated text file of sample information. Columns must have names. Sample/microarray ids should be listed under sample_name column. The cell-type (or 'group') of each sample should be listed under a column group_name.
dataset_name	Short, meaningful name for this dataset/experiment.
sample_name	Name of sample_sheet_table with sample ID
group_name	Name of sample_sheet_table with group/cell-type. Default = "group"
groups2test	An optional character vector specificing specific groups to check. By default (set to NA), all groups will be tested.
extra_factor_name	Optionally, an extra cross-group factor (as column name in sample_sheet_table) to include in the model used by limma. E.g. An individual/mouse id. Refer limma docs. Default = NA
pval_threshold	For reporting only, a p-value threshold. Default = 0.01

Details

Sometimes there are microarray studies measureing purified cell populations that would be measured together in a single-cell sequenicng experiment. E.g. comparing PBMC scRNA to FACs-sorted blood cell populations. This function will process microarray data with limma and format it for comparisions.

The microarray data used should consist of purified cell types from /emphone single study/experiment (due to batch effects). Ideally just those cell-types expected in the scRNAseq, but the method appears relatively robust to a few extra cell types.

Note that unlike the single-cell workflow there are no summarisedExperiment objects (they're not really comparable) - this function reads data and generates a table of within-dataset differentential expression contrasts in one step. Ie. equivalent to the output of contrast_each_group_to_the_rest.

Also, note that while downstream functions can accept the microarray-derived data as query datasets, its not really intended and assumptions might not hold (Generally, its known what got loaded onto a microarray!)

The (otherwise optional) 'limma' package must be installed to use this function.

Value

A tibble, the within-experiment de_table (differential expression table)

See Also

contrast_each_group_to_the_rest is the funciton that makes comparable output on the scRNAseq data (dataset_se objects).

Limma Limma package for differential expression.

Other Data loading functions: load_dataset_10Xdata, load_se_from_tables

Examples

contrast_each_group_to_the_rest_for_norm_ma_with_limma(
    norm_expression_table=demo_microarray_expr, 
    sample_sheet_table=demo_microarray_sample_sheet,
    dataset_name="DemoSimMicroarrayRef", 
    sample_name="cell_sample", group_name="group") 
    
## Not run:  
contrast_each_group_to_the_rest_for_norm_ma_with_limma(
   norm_expression_table, sample_sheet_table=samples_table, 
   dataset_name="Watkins2009PBMCs", extra_factor_name='description')

## End(Not run)

MonashBioinformaticsPlatform/celaref documentation built on June 5, 2019, 11:35 a.m.