load_geo_data: Load datasets from the GEO database

In miracum/clearly-sigident.preproc: Sigident Preprocessing

Load datasets from the GEO database

Description

Load datasets that are stored on the GEO databse

Usage

load_geo_data(
  studiesinfo,
  datadir = tempdir(),
  plotdir = "./plots/",
  idtype,
  viz_batch_boxp = TRUE,
  viz_batch_gpca = TRUE
)

Arguments

studiesinfo	A list that contains specifications on the studie's / studies' meta data.
datadir	A character string. Path to the folder to store the downloaded gene data sets (default: tempdir()).
plotdir	A character string. Path to the folder to store resulting plots. Default: "./plots/".
idtype	A character string. The type of ID used to name the genes. One of 'entrez' or 'affy' intended to use either entrez IDs or affy IDs. Caution: when using entrez IDs, missing and duplicated IDs are being removed!
viz_batch_boxp	A logical. Indicates, if batch effects and batch effect removals via histograms should be plotted (default: TRUE).
viz_batch_gpca	A logical. Indicates, if batch effects and batch effect removals via gPCA (gPCA::PCplot) should be plotted (default: TRUE).

Details

The functions writes objects to the global environment, including the expression sets of the studies specified in 'studiesinfo' - the here used study names (list keys) are used for the naming in the global environment. Furthermore, 'mergedset', a large expression set containing all studies, is also written to the global environment. Please note, that this set contains the raw expression data with batch effects. Bbatch effects are detected, removed and provided with the object 'mergeset', a matrix containing batch corrected expression data with genes in the rows and samples in the columns. 'sample_metadata' is a data.frame, which holds information on the samples which are included in the studies, including if they are a "Target" or a "Control". 'diagnosis' (a binary coding of the target variable) and 'batch' () are also provided to the global environment to be used with other functions of the 'sigident' R package.

References

https://gitlab.miracum.org/clearly/sigident

See Also

sigident

miracum/clearly-sigident.preproc documentation built on June 28, 2022, 3:17 p.m.