metaRanks: Meta Ranking of Transcriptome Genes

Description Usage Arguments Details Value Note Author(s)

View source: R/metaRank.R

Description

Combine several transcriptome or differential expression results to build one meta result

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
metaRanks(fnames, fids, weightset = rep(1, length(fileset)), 
		geneColumn = "GENE_ID", valueColumn="LOG_2_FOLD",
		pvalueColumn = "PVALUE", productColumn="PRODUCT", sep="\t",
		rank.average.FUN = sqrtmean, value.average.FUN = mean,
		keepIntergenics = FALSE,
		missingGenes = c("drop","fill","na"), missingValue = 0,
		naDropPercent = 0.5, nFDRsimulations=0)

metaRank.data.frames(df.list, weightset = rep(1, length(fileset)), 
		geneColumn = "GENE_ID", valueColumn="LOG_2_FOLD",
		pvalueColumn = "PVALUE", productColumn="PRODUCT", 
		rank.average.FUN = logmean, value.average.FUN = mean,
		missingGenes = c("drop","fill","na"), missingValue = 0,
		naDropPercent = 0.5)

metaRank2html( tbl, fileout = "metaRank.html", title = "", maxRows = 100,
		valueColumn = "LOG2FOLD", ...)

Arguments

fnames

character vector of full path names to existing transcript or DE files

fids

character vector of SampleIDs for each file

df.list

list of data frame objects, that are all tables of gene abundance information. This is just an alternative format to the file-based inputs.

weightset

numeric vector of weights for each file

geneColumn

column name that contains the GeneIDs

valueColumn

column name that contains the fold change or expression value

pvalueColumn

column name that contains the fold change or expression value

productColumn

column name that contains the fold change or expression value

rank.average.FUN

averaging function for combining the rank position of each gene

value.average.FUN

averaging function for combining the data value terms for each gene

keepIntergenics

logical, keep the explicit 'non-genes' or drop them from the result

missingGenes

mode for dealing with genes that are not found in every data set. drop drops the entire row/gene from the output. fill keeps the row/gene; and assigns the largest (worst) rank to every missing gene. na flags the gene's rank as NA, and then drops entire rows having genes missing from more than naDropPercent % of the datasets.

missingValue

value to use for the value term for missing genes

naDropPercent

minimum percentage of missing observations, to cause a gene to be dropped completely from the results

nFDRsimulations

number of trials of randomized trials of permuted ranks, to estimate the false positive rate

tbl

data frame result from metaRank functions, to be rendered as an HTML file

fileout

Full path name for the newly created HTML file.

title

Character string on length 1, to be used as header and title in the HTML file

maxRows

Maximum number of rows to include in the HTML file

...

Other arguments, passed to table2html

Details

This function combines several results files, while 'averaging' the results for each gene from all the separate files. All files are assumed to have the genes in a consistent ranked order, by either expression (transcripts), fold change (differential expression - DE), etc. For each gene, the average rank position and the average observed value are calculated, and that gene's rank position in each input file is reported. Final gene order is by average rank.

Value

A data frame that represents the average of the input files, along with each gene's rank position in every file.

Also, one last column FP_RATE of the simulated false discovery rates, which estimate how likely it was to see an average rank as high as that gene's, by chance.

Lastly, a correlation estimate of the pairwise Spearman rank correlation is measured and saved in the global workspace with the object name of metaRankCC.

Note

The false discovery rate step is not currently provided in the 'data.frame' version.

The final ranking can be sensitive to the choice of rank averaging function. The default of sqrtmean is a compromise between linear and geometric averaging.

Author(s)

Bob Morrison


robertdouglasmorrison/DuffyTools documentation built on Dec. 14, 2018, 3:05 p.m.