| Title: | Competitive Gene Sets Test with the Mann-Whitney-Wilcoxon Test |
|---|---|
| Description: | Friendly implementation of the Mann-Whitney-Wilcoxon test for competitive gene set enrichment analysis. |
| Authors: | Stefano Maria Pagnotta [aut, cre, cph]
|
| Maintainer: | Stefano Maria Pagnotta <[email protected]> |
| License: | GPL (>=3) |
| Version: | 1.0.1 |
| Built: | 2024-11-17 04:25:37 UTC |
| Source: | https://github.com/stefanomp/massivegst |
This function trims the table of results from massiveGST function retaining the rows with a logit2NES below the specified threshold.
cut_by_logit2NES(ttable, logit2NES_threshold = 0.58)cut_by_logit2NES(ttable, logit2NES_threshold = 0.58)
ttable |
a data frame of "mGST" class coming from massiveGST function. |
logit2NES_threshold |
a real value |
A data frame.
the functions cut_by_NES, cut_by_logit2NES, and cut_by_significance can be nested.
Stefano M. Pagnotta
Cerulo, Pagnotta (2021) doi:10.1101/2021.02.15.431228
massiveGST, cut_by_NES, cut_by_significance,
library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") head(ans) cut_by_logit2NES(ans) cut_by_logit2NES(cut_by_significance(ans)) plot(cut_by_logit2NES(ans))library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") head(ans) cut_by_logit2NES(ans) cut_by_logit2NES(cut_by_significance(ans)) plot(cut_by_logit2NES(ans))
This function trims the table of results from massiveGST function retaining the rows with a NES below the specified threshold.
cut_by_NES(ttable, NES_threshold = 0.6)cut_by_NES(ttable, NES_threshold = 0.6)
ttable |
a data frame of 'mGST' class coming from massiveGST function. |
NES_threshold |
a real value between 0.0 and 1. |
A data frame.
the functions cut_by_NES, cut_by_logit2NES, and cut_by_significance can be nested. In the case the test has alternative = 'two.sided', it is better to use cut_by_logit2NES for a symmetric trim of both directions.
Stefano M. Pagnotta
Cerulo, Pagnotta (2021) doi:10.1101/2021.02.15.431228
massiveGST, cut_by_logit2NES, cut_by_significance, summary.mGST, plot.mGST
library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "greater") head(ans) cut_by_NES(ans, NES_threshold = .65) summary(cut_by_NES(ans, NES_threshold = .65))library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "greater") head(ans) cut_by_NES(ans, NES_threshold = .65) summary(cut_by_NES(ans, NES_threshold = .65))
This function trims the table of results from massiveGST function according to the significance required.
cut_by_significance(ttable, level_of_significance = 0.05, where = c("BH.value", "bonferroni", "p.value") )cut_by_significance(ttable, level_of_significance = 0.05, where = c("BH.value", "bonferroni", "p.value") )
ttable |
a data frame of "mGST" class coming from massiveGST function. |
level_of_significance |
a real value between 0.0 and 1. |
where |
a character string specifying where the level_of_significance has to be applied to the output; must be one of "p.value", "BH.value" (default), and "bonferroni" |
BH.value is the adjustment of p-values according to Benijamini and Hockberg's method; B.value is the adjustment of p-values according to Bonferroni's method.
A data frame.
the functions cut_by_NES, cut_by_logit2NES, and cut_by_significance can be nested.
Stefano M. Pagnotta
Cerulo, Pagnotta (2021) doi:10.1101/2021.02.15.431228
massiveGST, cut_by_logit2NES, cut_by_NES, summary.mGST, plot.mGST
library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") head(ans) cut_by_significance(ans) cut_by_significance(ans, level_of_significance = 0.05, where = "p") cut_by_logit2NES(cut_by_significance(ans)) summary(cut_by_significance(ans, level_of_significance = 0.05, where = "bonferroni")) plot(cut_by_significance(ans, level_of_significance = 0.05, where = "bonferroni"))library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") head(ans) cut_by_significance(ans) cut_by_significance(ans, level_of_significance = 0.05, where = "p") cut_by_logit2NES(cut_by_significance(ans)) summary(cut_by_significance(ans, level_of_significance = 0.05, where = "bonferroni")) plot(cut_by_significance(ans, level_of_significance = 0.05, where = "bonferroni"))
Load a gene-profile from a txt file.
get_geneProfile(ffile)get_geneProfile(ffile)
ffile |
a character string or a list of a character pointing to a local file |
The txt file contains two columuns separated by a tabulation. The first column is the gene name ( or entrez, ensembl, etc); the second column are the numeric values associated with each gene. The profile do not need to be sorted.
As an example, see the file in /massiveGST/extdata/pre_ranked_list.txt
See the path in the example below.
A named list of numeric values.
Stefano M. Pagnotta
fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") fname geneProfile <- get_geneProfile(fname) class(geneProfile) head(geneProfile) tail(geneProfile)fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") fname geneProfile <- get_geneProfile(fname) class(geneProfile) head(geneProfile) tail(geneProfile)
Load the gene-sets collection from local gmt files
get_geneSets_from_local_files(ffiles)get_geneSets_from_local_files(ffiles)
ffiles |
a character string or a list of a character pointing to local files |
A vector list of gene-sets
Stefano M. Pagnotta
get_geneSets_from_msigdbr, write_geneSets_to_gmt
library(massiveGST) tmp <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") fname1 <- file.path(tempdir(), "h1.gmt") write_geneSets_to_gmt(tmp, fileName = fname1) fname2 <- file.path(tempdir(), "h2.gmt") write_geneSets_to_gmt(tmp, fileName = fname2) # getting one collection geneSets <- get_geneSets_from_local_files(fname1) length(geneSets) # getting two collections geneSets <- get_geneSets_from_local_files(c(fname1, fname2)) length(geneSets)library(massiveGST) tmp <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") fname1 <- file.path(tempdir(), "h1.gmt") write_geneSets_to_gmt(tmp, fileName = fname1) fname2 <- file.path(tempdir(), "h2.gmt") write_geneSets_to_gmt(tmp, fileName = fname2) # getting one collection geneSets <- get_geneSets_from_local_files(fname1) length(geneSets) # getting two collections geneSets <- get_geneSets_from_local_files(c(fname1, fname2)) length(geneSets)
This is a wrapper for extraction a gene-sets collection as a vector list to match the data structure for massiveGST function.
get_geneSets_from_msigdbr(category, what, subcategory = NULL, species = "Homo sapiens")get_geneSets_from_msigdbr(category, what, subcategory = NULL, species = "Homo sapiens")
category |
MSigDB collection abbreviation, such as H or C1. |
what |
a character string specifying the code representation of the genes; must be one of "gene_symbol", "entrez_gene", "ensembl_gene", "human_gene_symbol", "human_entrez_gene", "human_ensembl_gene"; |
subcategory |
MSigDB sub-collection abbreviation, such as CGP or BP; NULL (default) |
species |
Species name, such as 'Homo sapiens' or 'Mus musculus'. |
A vector list of gene-sets
Stefano M. Pagnotta
library(massiveGST) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") class(geneSets) head(geneSets, 3)library(massiveGST) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") class(geneSets) head(geneSets, 3)
Perform a competitive gene set enrichment analysis by applying the Mann-Withney-Wilcoxon test.
massiveGST(gene_profile, gene_sets, cols_to_remove = NULL, alternative = c("two.sided", "less", "greater") )massiveGST(gene_profile, gene_sets, cols_to_remove = NULL, alternative = c("two.sided", "less", "greater") )
gene_profile |
a named list of values; the names have to match the names fo genes in the gene-set. |
gene_sets |
a character vector of gene-sets |
cols_to_remove |
a list of colnames to eventually remove from the output |
alternative |
a character string specifying the alternative hypothesis of the MWW test; must be one of "two.sided" (default), "greater" or "less". |
A data frame with columns
size |
Original size of the gene-set |
actualSize |
Size of the gene-set after the match with the gene-profile |
NES |
(Normalized Enrichment Score) the strength of the association of the gene-set with the gene profile; also the percentile rank of the gene-set in the universe of the genes ouside the gene-set. |
odd |
odd transformation of the NES |
logit2NES |
logit transformation of the NES |
abs_logit2NES |
absolute value of the logit2NES in the case of "two.sided" alternative |
p.value |
p-values associated with the gene-set |
BH.value |
Benijamini and Hockberg adjustment of the p.values |
B.value |
Bonferroni adjustment of the p.values |
relevance |
marginal ordering of the table |
Stefano M. Pagnotta
Cerulo, Pagnotta (2021) doi:10.1101/2021.02.15.431228
summary.mGST, plot.mGST, cut_by_logit2NES, cut_by_NES, cut_by_significance
library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") anslibrary(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") ans
This function displays the enrichment analysis results both as a bar-plot and a network of gene-sets.
## S3 method for class 'mGST' plot(x, gene_sets = NULL, order_by = "logit2NES", top = 30, eps = 0.25, as.network = FALSE, similarity_threshold = 1/3, manipulation = FALSE, autoResize = TRUE, ... )## S3 method for class 'mGST' plot(x, gene_sets = NULL, order_by = "logit2NES", top = 30, eps = 0.25, as.network = FALSE, similarity_threshold = 1/3, manipulation = FALSE, autoResize = TRUE, ... )
x |
a data structure coming from the massiveGST function |
gene_sets |
a character vector of gene-sets; mandatory for the network display |
order_by |
a character string specifying whick should be the ordering in the bar-plot; must be one of "relevance", "NES", "logit2NES" (default), "p.value", "BH.value", and "bonferroni". These are the same options of summary.mGST |
top |
an integer value controlling how many gene-sets have to be displaued in the bar-plot; top = 30 (default) |
as.network |
a logical value to switch to a network display; as.network = FALSE (default) |
similarity_threshold |
a real value to cut the similarities between gene-stes below this value; similarity_threshold = 1/3 (default) |
eps |
a real value between 0.0 and 1.0 controlling the contribution of the Jaccard and overlap similaties to their convex combination; eps = 0.25 (default), see details. |
manipulation |
a logical value allowing to manipulate the network; manipulation = FALSE (default); see visOptions |
autoResize |
a logical value allowing to resize the network; resize = TRUE (default); see visOptions |
... |
other graphical parameters |
This function display the results of enrichment analysis both as a bar-plot and a network.
The network rendering is with the visNetwork package.
The similarity between the gene-set is computed a convex combination of the Jaccard and overlap similarities. See the reference for further details.
In the case of network display, an object from the visNetwork package.
Stefano M. Pagnotta
Cerulo, Pagnotta (2021) doi:10.1101/2021.02.15.431228
massiveGST, visNetwork, visOptions
library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") # to get the bar-plot plot(cut_by_significance(ans, level_of_significance = 0.01)) # to get the network of the gene-sets plot(cut_by_significance(ans, level_of_significance = 0.01), gene_sets = geneSets, as.network = TRUE)library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") # to get the bar-plot plot(cut_by_significance(ans, level_of_significance = 0.01)) # to get the network of the gene-sets plot(cut_by_significance(ans, level_of_significance = 0.01), gene_sets = geneSets, as.network = TRUE)
This gene-profile comes from the paper in reference. It compares 9 FGFR3-TACC3 fusion positive samples versus 535 other samples in the GBM study from TCGA (Agilent platform).
Stefano M. Pagnotta
Frattini et al. "A metabolic function of FGFR3-TACC3 gene fusions in cancer" Nature volume 553, 2018 doi:10.1038/nature25171
Save the data frame coming from the massiveGST function as tab-separeted value.
save_as_tsv(x, file_name = "massiveGST.tsv", sep = "\t", ...)save_as_tsv(x, file_name = "massiveGST.tsv", sep = "\t", ...)
x |
a data frame of "mGST" class coming from massiveGST function. |
file_name |
a character value ("massiveGST.tsv" as default) |
sep |
a character value |
... |
Arguments to be passed to methods |
No return value.
Stefano M. Pagnotta
library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") # save the results fname <- file.path(tempdir(), "massiveGST_results.tsv") save_as_tsv(ans, file_name = fname)library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") # save the results fname <- file.path(tempdir(), "massiveGST_results.tsv") save_as_tsv(ans, file_name = fname)
Save the data frame coming from the massiveGST function as Excel 2003 (XLS) or Excel 2007 (XLSX) files
save_as_xls(x, file_name = "massiveGST.xls", ...)save_as_xls(x, file_name = "massiveGST.xls", ...)
x |
a data frame of "mGST" class coming from massiveGST function. |
file_name |
a character value ("massiveGST.xls" as default) |
... |
Arguments to be passed to methods |
No return value.
Stefano M. Pagnotta
library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") # save the results fname <- file.path(tempdir(), "massiveGST_results.xls") save_as_xls(ans, file_name = fname)library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") # save the results fname <- file.path(tempdir(), "massiveGST_results.xls") save_as_xls(ans, file_name = fname)
This method handles the result of massiveGST function, to provide views of the table.
## S3 method for class 'mGST' summary(object, cols_to_remove = "link", order_by = c("relevance", "NES", "logit2NES", "p.value", "BH.value", "bonferroni"), top = NULL, as.formattable = FALSE, ... )## S3 method for class 'mGST' summary(object, cols_to_remove = "link", order_by = c("relevance", "NES", "logit2NES", "p.value", "BH.value", "bonferroni"), top = NULL, as.formattable = FALSE, ... )
object |
a data structure coming from the massiveGST function |
cols_to_remove |
A character list of the columns to remove from the output. |
order_by |
a character string specifying which marginal ordering has to be applied to the output; must be one of "relevance" (default), "NES", "logit2NES", "p.value", "BH.value", and "bonferroni" |
top |
an integer to trim the table to the first 'top' rows. |
as.formattable |
a logical value (default = FALSE) to provide a formatted output with the help of formattable package. |
... |
Arguments to be passed to methods |
A data frame.
Stefano M. Pagnotta
library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") summary(ans) summary(ans, as.formattable = TRUE, order_by = "NES", top = 10)library(massiveGST) # get the gene profile fname <- system.file("extdata", package="massiveGST") fname <- file.path(fname, "pre_ranked_list.txt") geneProfile <- get_geneProfile(fname) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # run the function ans <- massiveGST(geneProfile, geneSets, alternative = "two.sided") summary(ans) summary(ans, as.formattable = TRUE, order_by = "NES", top = 10)
Write a collection of gene sets as arranged in this package in a gmt file format.
write_geneSets_to_gmt(gs, fileName)write_geneSets_to_gmt(gs, fileName)
gs |
a character vector of gene-sets |
fileName |
a character value; "gene_sets.gmt" (default) |
No return value.
Stefano M. Pagnotta
get_geneSets_from_msigdbr, get_geneSets_from_local_files
library(massiveGST) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # save the gene-sets fname <- file.path(tempdir(), "hallmarks.gmt") write_geneSets_to_gmt(geneSets, fileName = fname)library(massiveGST) # get the gene-sets geneSets <- get_geneSets_from_msigdbr(category = "H", what = "gene_symbol") # save the gene-sets fname <- file.path(tempdir(), "hallmarks.gmt") write_geneSets_to_gmt(geneSets, fileName = fname)