GCSscore: Main GCS-score Function

In harrisgm/GCSscore: GCSscore: an R package for microarray analysis for Affymetrix/Thermo Fisher arrays

Main GCS-score Function

Description

The main function used to call and run the GCS-score algorithm.

Usage

GCSscore(celFile1 = NULL, celFile2 = NULL, celTable = NULL,celTab.names = FALSE,
typeFilter = 0, method = 1, rm.outmask = FALSE, SF1 = NULL, SF2 = NULL, fileout = FALSE, 
gzip = FALSE, verbose = FALSE)

Arguments

celFile1	If a one comparison run is desired, enter the filename and path to the 1st Affymetrix CEL file
celFile2	If a one comparison run is desired, enter the filename and path to the 2nd Affymetrix CEL file
celTable	If a batch run is desired, enter the filename and path to the CSV file containing the batch information
celTab.names	If set to TRUE, then the GCS-score batch output is assigned the user-designated name (specified in the first column of the celTable CSV file (see examples))
typeFilter	If set to 0, all available probe types are included in the calculation and normalization of the GCS-score values. If set to 1, only probes well-annotated probe_ids (from BioConductor .db packages) are included in the calculation and normalization of the GCS-score output
method	This determines the method used to group and tally the probes_ids when calculating GCS-scores. For Whole Transcriptome arrays, for gene-level (transcript_cluster_id-based) analysis, set method = 1, and for exon-level (probeset_id-based) analysis, set method = 2. For the older generation arrays (3' IVT-style), if a PM-MM based background correction is desired, set method = 1 (PM-MM gives identical results to the original sscore package). If a GC-content based background correction is desired on the 3' IVT arrays, set method = 2
rm.outmask	If set to TRUE, then probes that are flagged as MASKED or OUTLIER in either CEL file 1 or CEL file 2 will be removed from the analysis. If set to FALSE, these probes are not filtered out and will be used in the GCS-score calculation
SF1	Input a pre-determined Scaling Factor (SF) for the 1st CEL file
SF2	Input a pre-determined Scaling Factor (SF) for the 2nd CEL file
fileout	Determines if the resulting GCS-score output is written to disk in a CSV format following the completion of the function.
gzip	If set to TRUE, the GCSscore output that is written to disk is compressed. This could prove useful if a large number runs are input using the batch submission
verbose	If set to TRUE, more information will be printed to the console during while the algorithm is running

Details

The input accepts individual CEL files or reads in a CSV file for batch runs. The user also inputs parameters to determine the method used by the GCS-score algorithm to group and tally the individual probes on a given array.

Value

An ExpressionSet object with GCS-score values for the probe groupings (determined by the method argument) and the relevant annotation informtaion

Examples

if (length(list.files(path = ".", pattern = "*.CEL")) != 0){

######################## Single run example ###########################

# get the path to example CEL files provided with package:
celpath1 <- system.file("extdata/","MN_2_3.CEL", package = "GCSscore")
celpath2 <- system.file("extdata/","MN_4_1.CEL", package = "GCSscore")

# run GCSscore() function directly on the two .CEL files above:
GCSs.single <- GCSscore::GCSscore(celFile1 = celpath1, celFile2 = celpath2)

# convert GCSscore single-run from ExpressionSet to data.table:
GCSs.single.dt <- 
  data.table::as.data.table(cbind(GCSs.single@featureData@data,
                                  GCSs.single@assayData[["exprs"]]))

# show all column names included in the output:
colnames(GCSs.single.dt)

# show simplified output of select columns and rows:
GCSs.single.dt[10000:10005,
              c("transcriptclusterid","symbol",
                "ref_id","Sscore")]

######################## batch run example ############################

# get the path to example batch (.csv) file provided with package:
celtab_path <- system.file("extdata",
                           "GCSs_batch_ex.csv", 
                           package = "GCSscore")

# read in the .CSV file using fread():
celtab <- data.table::fread(celtab_path)

# view structure of 'celTable' input:
celtab

# add the path to the sample CEL files to the batch input:
#   NOTE: this step is not necessary if the .CEL files 
#         are in the working directory:

path <- system.file("extdata", package = "GCSscore")
celtab$CelFile1 <- celtab[,paste(path,CelFile1,sep="/")]
celtab$CelFile2 <- celtab[,paste(path,CelFile2,sep="/")]

# run GCSscore function on the batch input:
GCSs.batch <- GCSscore::GCSscore(celTable = celtab, celTab.names = TRUE)

# convert GCS-score output from 'ExpressionSet' to 'data.table':
GCSs.batch.dt <-
  data.table::as.data.table(cbind(GCSs.batch@featureData@data,
                                  GCSs.batch@assayData[["exprs"]]))
                                  
# show all columns included in the output:
colnames(GCSs.batch.dt)

# show simplified output of GCSscore batch example:
GCSs.batch.dt[10000:10005,
              c("transcriptclusterid","symbol",
                "example01","example02","example03")]
}

harrisgm/GCSscore documentation built on Jan. 1, 2023, 12:04 a.m.