detect_SL: Prediction of Synthetic Lethality

In TomKellyGenetics/slipt: Synthetic Lethal Interaction Prediction Tool (SLIPT) from Cancer Genomics Data

Description

Statistical detection of expression signals consistent with synthetic lethal interaction. Based on the Chi-Square test with built-in multiple testing corrections.

Usage

detect_SL(query, datasetx)

Arguments

query	string. Name of a gene to test all others against. Note this gene must be contained within the dataset, matching a column name of datasetx.
datasetx	Dataset input to be formatted. Input data is: samples (rows) x genes (columns) as formatted by prep_data_for_SL. Output data is a table of genes (rows) x test statistics and outputs (columns), in order of most significantly supported candidate interacting parnter, this tabular format is human-readable is handled by count_of_SL or table_of_SL for built-in summary statistics.

Examples

#prepare data
 data <- c()
for(i in 1:100){
  data <- cbind(data, rnorm(1000))
}
rm(i)
rownames(data) <- paste("gene", 1:1000)
colnames(data) <- paste("sample", 1:100)
partitioned_data <- prep_data_for_SL(data, n = 3)

#run SLIPT
sl_table <- detect_SL("gene 1", partitioned_data)
dim(sl_table)
head(dim(sl_table))

sl_table <- detect_SL("gene 1", prep_data_for_SL(data))
dim(sl_table)
head(dim(sl_table))

TomKellyGenetics/slipt documentation built on May 8, 2019, 9:28 a.m.