GSNA, Version v0.1.6.7

This version contains numerous enhancements and bug fixes for gsnMergePathways(), gsnSubnetsDotPlot(), yassifyPathways(), termSummary(), and gsnPathways<-().

GSNA version v0.1.6.4

This version adds tests for the gsnPathways() and buildGeneSetNetworkKappa() functions, and contains fixes for gsnSubnetsDotPlot().

GSNA, Version v0.1.6.1

This version adds several new features, including:

• gsnSubnetsDotPlot(): a function for generating a dot plots for subnets/gene set clusters, such that gene sets within a subnet are grouped together on the y-axis, whereas the x-axis represents confidence metric from a pathways analysis. Additional aesthetics such as color and dot size are configurable.
• termSummary(): a function for taking a set of descriptive gene set terms/names and extracting common terms as discriptors of gene set subnets.
• gsnPathways() function to retrieve and assign columns within the pathways data.frame.
• Updates to the vignette.

GSNA, Version v0.1.5

• In previous versions of the GSEA package, sample data included in the vignette contained gene sets derived from the academic version of the dorothea package, and therefore may not be used for commercial purposes. In this version, the GSEA example in the vignette and the corresponding gene sets from the dorothea package used as example data in the vignette were updated to include only genes sets and genes from the non-academic version of the dorothea package, so as to be compatible with commercial use.
• Includes numerous corrections to the vignette.
• Additional bug fixes and tests are incorporated in this release version.

GSNA version v0.1.4.7

• Fixes a serious bug in lfisher_cpp() that intermittantly produced erroneously significant log Fisher p-values with large backgrounds/numbers of total genes. This bug, which also affected GSNA::buildGeneSetNetworkSTLF(), GSNA::buildGeneSetNetworkLF() and GSNA::scoreLFMatrix_C(), was due to an interaction between a rounding error in the summation of partial p-values with a programming logic issue. Also, added tests to tests/test-lfisher_cpp.R to detect this problem.
• Fixed numeric overrun problem with the lse() function that led to incorrect calculation of log harmonic means with the private lhm() function, which resulted in incorrect log harmonic mean P values being calculated by gsnSubnetSummary(). This bug was a contributor to incorrect calculation of the log of harmonic means of within subnet single-tail log-Fisher values.
• Fixed a sorting issue that also contributed to incorrect calculation of log harmonic mean within-subnet single tail Fisher p-values.
• Updated yassifyPathways() so that identifiers containing colons such as GO terms (e.g. "GO:0045824") can be used with the url_map_by_words_list argument. Currently, word characters "\\w" and colons ":" can be used.
• Fixed problem with gsnImportGenericPathways() that caused explicitly specified values of stat_col to be ignored.

GSNA, Version v0.1.4.3

• Includes the ./man directory and .Rd files as part of the GitHub repository, so that installations from GitHub via devtools::install_github() will now include the manual pages, without the need to run devtools::document() or roxygen2::roxygenize().
• Adds tests for gsIntersectCounts() & lfisher_cpp(), and silences the messages for gsnAddPatheaysData() and gsnImportGenericPathways().
• Includes the ./man directory and .Rd files as part of the GitHub repository, so that installations from GitHub via devtools::install_github() will now include the manual pages, without the need to run devtools::document() or roxygen2::roxygenize().
• This is intended to be the last version with dependency on the obsolete raster library.

GSNA, Version v0.1.4.1

This is the first public release of GSNA on CRAN.

GSNA stands for Gene Set Network Analysis, and it is a set of tools for creating networks of gene sets, inferring clusters of functionally-related gene sets based on similarity statistics, and visualization of the results. This package simplifies and accelerates interpretation of pathways analysis data sets. It is designed to work in tandem with standard pathways analysis methods, such as the 'GSEA' program (Gene Set Enrichment Analysis), CERNO (Coincident Extreme Ranks in Numerical Observations, implemented in the 'tmod' package) and others. Inputs to 'GSNA' are the outputs of pathways analysis methods: a list of gene sets (or "modules"), pathways or GO-terms with associated p-values. Since pathways analysis methods may be used to analyze many different types of data including transcriptomic, epigenetic, and high-throughput screen data sets, the 'GSNA' pipeline is applicable to these data as well. The use of 'GSNA' has been described in the following papers:

• Collins DR, Urbach JM, Racenet ZJ, Arshad U, Power KA, Newman RM, et al. (2021) doi:10.1016/j.immuni.2021.08.007
• Collins DR, Hitschfel J, Urbach JM, Mylvaganam GH, Ly NL, Arshad U, et al. (2023) doi:10.1126/sciimmunol.ade5872.