Several gene set databases are publicly available and can be used for
gene set analysis as part of the screening of bulk data for hypothesis
generation. The Hallmark gene set database [1] provides a well-curated
gene set list of biological states which can be used to obtain an
overall characterization of the data. Other Human Molecular Signatures
Database (MSigDB) collections including gene ontology pathways
(Biological Process, Cellular Component, Molecular Function),
immunologic signature gene sets or regulatory target gene sets can be
employed for roastgsa analysis to identify the most
relevant expression changes between experimental conditions for highly
specific biological functions. These broad collections and
sub-collections can be downloaded through https://www.gsea-msigdb.org/gsea/msigdb/collections.jsp.
Other public collections that can be employed for battery testing such
as the KEGG pathway database (https://www.genome.jp/kegg/pathway.html) [2] or reactome
(https://reactome.org/)
[3] provide a broad representation of gene sets for human diseases,
metabolism or cellular processes among others.
These gene set collections can be considered for
roastgsa either by (1) R loading the gene sets in a
list object, each element containing the gene identifiers
for the testing set or (2) saving a .gmt file with the
whole collection in a specific folder, i.e.,
We download publicly available arrays from GEO, accession ‘GSE145603’, which contain LGR5 and POLI double tumor cell populations in colorectal cancer [4]
# DO NOT RUN
# library(GEOquery)
# data <- getGEO('GSE145603')
# normdata <- (data[[1]])
# pd <- pData(normdata)
# pd$group_LGR5 <- pd[["lgr5:ch1"]]We are interested in screening the hallmarks with the largests
changes between LGR5 negative and LGR5 positive samples. The hallmark
gene sets are stored in gspath with the file name specified
in gsetsel (see specifications above).
The formula, design matrix and the corresponding contrast can be obtained as follows
# DO NOT RUN
# form <- "~ -1 + group_LGR5"
# design <- model.matrix(as.formula(form),pd)
# cont.mat <- data.frame(Lgr5.high_Lgr5.neg = c(1,-1))
# rownames(cont.mat) <- colnames(design)In microarrays, the expression values for each gene can be measured in several probesets. To perform gene set analysis, we select the probeset with maximum variability for every gene:
# DO NOT RUN
# mads <- apply(exprs(normdata), 1, mad)
# gu <- strsplit(fData(normdata)[["Gene Symbol"]], split=' \\/\\/\\/ ')
# names(gu) <- rownames(fData(normdata))
# gu <- gu[sapply(gu, length)==1]
# gu <- gu[gu!='' & !is.na(gu) & gu!='---']
# ps <- rep(names(gu), sapply(gu, length))
# gs <- unlist(gu)
# pss <- tapply(ps, gs, function(o) names(which.max(mads[o])))
# psgen.mvar <- pssThe roastgsa function is used for competitive gene set
analysis testing:
# DO NOT RUN
# roast1 <- roastgsa(exprs(normdata), form = form, covar = pd,
# psel = psgen.mvar, contrast = cont.mat[, 1],
# gspath = gspath, gsetsel = gsetsel, nrot = 1000,
# mccores = 7, set.statistic = "maxmean")
#
# print(roast1)
#Summary tables can be presented following the
roastgsa::htmlrgsa documentation.
## R version 4.5.1 (2025-06-13)
## Platform: x86_64-pc-linux-gnu
## Running under: Ubuntu 24.04.3 LTS
##
## Matrix products: default
## BLAS: /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3
## LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so; LAPACK version 3.12.0
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_US.UTF-8 LC_COLLATE=C
## [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
## [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## time zone: Etc/UTC
## tzcode source: system (glibc)
##
## attached base packages:
## [1] stats4 stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] DESeq2_1.51.0 SummarizedExperiment_1.41.0
## [3] Biobase_2.71.0 MatrixGenerics_1.23.0
## [5] matrixStats_1.5.0 GenomicRanges_1.63.0
## [7] Seqinfo_1.1.0 IRanges_2.45.0
## [9] S4Vectors_0.49.0 BiocGenerics_0.57.0
## [11] generics_0.1.4 roastgsa_1.9.0
## [13] knitr_1.50 BiocStyle_2.39.0
##
## loaded via a namespace (and not attached):
## [1] gtable_0.3.6 xfun_0.54 bslib_0.9.0
## [4] ggplot2_4.0.0 caTools_1.18.3 lattice_0.22-7
## [7] vctrs_0.6.5 tools_4.5.1 bitops_1.0-9
## [10] parallel_4.5.1 tibble_3.3.0 pkgconfig_2.0.3
## [13] Matrix_1.7-4 KernSmooth_2.23-26 RColorBrewer_1.1-3
## [16] S7_0.2.0 lifecycle_1.0.4 compiler_4.5.1
## [19] farver_2.1.2 gplots_3.2.0 statmod_1.5.1
## [22] codetools_0.2-20 htmltools_0.5.8.1 sys_3.4.3
## [25] buildtools_1.0.0 sass_0.4.10 yaml_2.3.10
## [28] pillar_1.11.1 jquerylib_0.1.4 BiocParallel_1.45.0
## [31] cachem_1.1.0 DelayedArray_0.37.0 limma_3.67.0
## [34] abind_1.4-8 gtools_3.9.5 tidyselect_1.2.1
## [37] locfit_1.5-9.12 digest_0.6.37 dplyr_1.1.4
## [40] labeling_0.4.3 maketools_1.3.2 fastmap_1.2.0
## [43] grid_4.5.1 cli_3.6.5 SparseArray_1.11.1
## [46] magrittr_2.0.4 S4Arrays_1.11.0 withr_3.0.2
## [49] scales_1.4.0 rmarkdown_2.30 XVector_0.51.0
## [52] evaluate_1.0.5 rlang_1.1.6 Rcpp_1.1.0
## [55] glue_1.8.0 BiocManager_1.30.26 jsonlite_2.0.0
## [58] R6_2.6.1[1] A. Liberzon, C. Birger, H. Thorvaldsdottir, M. Ghandi, J. P. Mesirov, and P. Tamayo. The Molecular Signatures Database Hallmark Gene Set Collection. Cell Systems, 1(6):417-425, 2015.
[2] M. Kanehisa et al. KEGG as a reference resource for gene and protein annotation, Nucleic Acids Research, Volume 44, Issue D1, 4 January 2016, D457–D462, https://doi.org/10.1093/nar/gkv1070
[3] M. Gillespie et al. The reactome pathway knowledgebase 2022, Nucleic Acids Research, 2021;, gkab1028, https://doi.org/10.1093/nar/gkab1028
[4] Morral C, Stanisavljevic J, Hernando-Momblona X, et al. Zonation of Ribosomal DNA Transcription Defines a Stem Cell Hierarchy in Colorectal Cancer. Cell Stem Cell. 2020;26(6):845-861.e12. doi:10.1016/j.stem.2020.04.012