dea_ebseq: Run EBSeq gene Differential Expression Analysis (DEA).

In Facottons/DOAGDC: A Package to Download, Organize and Analyze Genomic Data Commons (GDC) Data

Description

Run EBSeq gene Differential Expression Analysis (DEA).

Usage

dea_ebseq(
  name,
  work_dir,
  env,
  tumor,
  group_gen,
  clinical_pair,
  pair_name = "G2_over_G1",
  rounds = 7,
  norm_type = "QuantileNorm",
  ebtest_qtrm = 0.75,
  ebtest_qtrm_cut = 10,
  p_cutoff = 0.05,
  fdr_cutoff = 0.05,
  fc_cutoff = 2,
  width = 2000,
  height = 1500,
  res = 300,
  unit = "px",
  image_format = "png",
  bullard_quantile = 0.75
)

Arguments

name	A character string indicating the desired values to be used in next analysis. For instance, "HIF3A" in the legacy gene expression matrix, "mir-1307" in the miRNA quantification matrix, or "HER2" in the protein quantification matrix.
work_dir	A character string specifying the path to work directory.
env	A character string containing the environment name that should be used. If none has been set yet, the function will create one in global environment following the standard criteria: 'tumor_data_base_data_type_tumor_data' or 'tumor_data_base_data_type_both_data' (for tumor and not tumor data in separated matrices).
tumor	A character string contaning one of the 33 tumors available in the TCGA project. For instance, the "BRCA" stands for breast cancer.
group_gen	A character string of the groups generation function: "mclust" - groups_identification_mclust(); "CoxHR" - groups_identification_coxHR(); "clinical" groups_identification_clinical().
clinical_pair	A character string containing one of the group pairs selected after statistical analysis runned in clinical_terms() function.
pair_name	A character string indicating which condition name should be used. When there are only two groups the default is "G2_over_G1".
rounds	Numerical value indicating the number of iterations. It is recommended to check the Alpha and Beta convergence plots in output and adjust this value until the hyper-parameter estimations converged. The default is 7.
norm_type	A character string indicating which EBSeq normalization factors type should be used in the analysis "QuantileNorm" or "MedianNorm". 'The default is "all".
ebtest_qtrm, ebtest_qtrm_cut	Numerical value. It is removed from the analysis genes with ebtest_qtrm th quantile < = ebtest_qtrm_cut. More details in EBSeq EBTest page. The default is ebtest_qtrm = 0.75 and ebtest_qtrm_cut = 10.
p_cutoff	Numerical value indicating the maximum value for p-values. The default is 0.05.
fdr_cutoff	Numerical value indicating the maximum value for FDR values. The default is 0.05.
fc_cutoff	Numerical value indicating the maximum value for Fold Change '(FC). The default is 2.
width	Graphical parameters. See par for more details. As default width = 2000, height = 1500, res = 300 and unit = "px".
height	Graphical parameters. See par for more details. As default width = 2000, height = 1500, res = 300 and unit = "px".
res	Graphical parameters. See par for more details. As default width = 2000, height = 1500, res = 300 and unit = "px".
unit	Graphical parameters. See par for more details. As default width = 2000, height = 1500, res = 300 and unit = "px".
image_format	A character string indicating which image_format will be used. It could be "png" or "svg". The only unit available in "svg" is inches ('in'). The default is "png".
bullard_quantile	Numerical value indicating the quantile for the Bullard's normalization. The default is 0.75.

Value

A matrix with DE genes in row and statistical values in columns.

Examples

library(DOAGDC)

# data already downloaded using the 'download_gdc' function
concatenate_expression("gene",
    name = "HIF3A",
    data_base = "legacy",
    tumor = "CHOL",
    work_dir = "~/Desktop"
)

# separating gene HIF3A expression data patients in two groups
groups_identification_mclust("gene", 2,
    name = "HIF3A",
    modelName = "E",
    env = CHOL_LEGACY_gene_tumor_data,
    tumor = "CHOL"
)

# load not normalized data
concatenate_expression("gene",
    normalization = FALSE,
    name = "HIF3A",
    data_base = "legacy",
    tumor = "CHOL",
    env = CHOL_LEGACY_gene_tumor_data,
    work_dir = "~/Desktop"
)

# start DE analysis
# considering concatenate_expression and groups_identification already runned
dea_ebseq(
    pair_name = "G2_over_G1",
    rounds = 7,
    name = "HIF3A",
    env = CHOL_LEGACY_gene_tumor_data
)

Facottons/DOAGDC documentation built on April 7, 2020, 3:17 a.m.