do_elbow_rnaseq {ELBOW} | R Documentation |
Performs the ELBOW fold change test on an CountDataSet from DESeq object. This is a wrapper class to help integrate the ELBOW method into Bioconductor. followed tutorial from: http://cgrlucb.wikispaces.com/Spring+2012+DESeq+Tutorial
do_elbow_rnaseq(rnaSeq)
rnaSeq |
is the CountDataSet object to analyze. |
a matrix specified as follows
columns — (1) “up_limit”, the upper ELBOW fold-change cut-off value; (2) “low_limit”, the lower ELBOW fold-change cut-off value
rows — one row per sample, specified by the parameter “columns.”
# install the DESeq libraries #source("http://www.bioconductor.org/biocLite.R") #biocLite("DESeq") ## download the table library("DESeq") # the following bam file dataset was obtained from: # http://cgrlucb.wikispaces.com/file/view/yeast_sample_data.txt # it has been downloaded into this package for speed convenience. filename <- system.file("extdata", "yeast_sample_data.txt", package = "ELBOW") count_table <- read.table(filename, header=TRUE, sep="\t", row.names=1) expt_design <- data.frame(row.names = colnames(count_table), condition = c("WE","WE","M","M","M")) conditions = expt_design$condition data <- newCountDataSet(count_table, conditions) data <- estimateSizeFactors(data) data <- as(data, "CountDataSet") ## data <- estimateVarianceFunctions(data) data <- estimateDispersions(data) # this next step is essential, but it takes a long time... # so, just like a good cooking show we will skip this step # and load a finished version. #results <- nbinomTest(data, "M", "WE") # The below two code lines load a copy of the above dataset # which has already been processed by: # results <- nbinomTest(data, "M", "WE") # For your own real data, you must use: # results <- nbinomTest(data, "M", "WE")' # Instead of the two lines below: data(yeast_nbinomTest_results, package="ELBOW") results <- yeast_nbinomTest_results # obtain the elbow limit for the dataset # the final step in the analysis pipeline do_elbow_rnaseq(results)