PADi: PADi

In huangwb8/GSClassifier: Gene Signature Classifier

PADi

Description

PAD for individuals

Usage

PADi(
  X,
  geneid = "ensembl",
  matchmode = c("fix", "free")[1],
  cancer.type = "GC",
  version = c("20200110", "20220916")[1],
  numCores = 0,
  verbose = T
)

Arguments

X	a numeric RNA expression vector or matrix with gene names
geneid	type of gene id in X. One of 'symbol', 'entrez' and 'ensembl'.
matchmode	If your genes belong to ENSEMBL, SYMBOL or ENTREZ, you should set matchmode = 'fix', in which all genes would be aligned to ENSEMBL. 'matchmode = free' is a legacy feature, which means that you have to convert your genes into the same annotation type of the pre-trained GSClassifier model.
cancer.type	The type of cancer
version	model version
numCores	No. of CPU core
verbose	whether report messages

Details

The current software only supports parameters: cancer.type = "GC", version = "20200110". More cancer types or model version were under developing.

Value

A data frame containing personalized PAD subtypes

Author(s)

Weibin Huang<[e-mail]>

See Also

callEnsemble; parCallEnsemble.

Examples

# Package
if (!requireNamespace("GSClassifier", quietly = TRUE)){
  devtools::install_github("huangwb8/GSClassifier")
}
library(GSClassifier)

# Load test data of RNA-Seq expression
testData <- readRDS(system.file("extdata", "testData.rds", package = "GSClassifier"))[['Kim2018_3']]
dim(testData) # 19118 3

# Data formats (Different data but the same function and usage)

# 1.Mutiple data
X = testData

# 2.Single data
# X <- testData[,1]; names(X) <- rownames(testData)
# X <- as.matrix(testData[,1]); rownames(X) <- rownames(testData)

## Method1: use a general function called 'callEnsemble'
res_padi <- callEnsemble(
  X = X,
  ens = NULL,
  geneAnnotation = NULL,
  geneSet = NULL,
  scaller = NULL,
  geneid = "ensembl",
  matchmode = 'fix',
  subtype = "PAD.train_20200110",
  verbose = F
)

## Method2: use a specific function called 'PADi'
res_padi <- PADi(X = X, verbose = F)

## Method3: Parallel strategy for lots of samples (empirically >50) to save #' time. Not Run for small cohorts.
# res_padi <- parCallEnsemble(
#   X = X,
#   ens = NULL,
#   geneAnnotation = NULL,
#   geneSet = NULL,
#   scaller = NULL,
#   geneid = 'ensembl',
#   matchmode = 'fix',
#   subtype = 'PAD.train_20200110',
#   verbose = T,
#   numCores = 2)
# res_padi <- PADi(X = X, verbose = F, numCores = 4)

huangwb8/GSClassifier documentation built on July 12, 2024, 5:10 p.m.