Skip to content

Latest commit

 

History

History
92 lines (59 loc) · 3.37 KB

NEWS.md

File metadata and controls

92 lines (59 loc) · 3.37 KB
Raw

News about the tcgsaseq R package

Main changes in Version 2.0.1 (2020-09-09)

  • renaming varseq() function into varcompseq (following trademark infringement notification from Golden Helix, Inc)
  • indicating soft-deprecation of tcgsaseqin favor of the dearseq Bioconductor package

Main changes in Version 2.0.0 (2019-08-02):

  • transform = TRUE is now the default
  • loc-lin weights are now computed on unconditional (transformed) counts. They should be more robust this way to outlier genes

Main changes in Version 1.9.0 (2019-07-18):

  • (partial) parallel support for permutations
  • progressbar for permutations in interactive mode
  • bug fix for voom-like weights

Main changes in Version 1.8.1 (2018-12-06):

  • perm_pe function for exact p-values and accurate FDR control when using permutation test. The column names of the $pvals output are back to being the same regardless of the test chosen being asympotic or permutation
  • dsFDR is not used anymore

Main changes in Version 1.8.0 (2018-11-23):

  • dsFDR function for accurate discrete FDR control when using permutation test. This changes the column names of the $pvals output when which_test = permutations

Main changes in Version 1.7.2 (2018-07-18):

  • faster permutations

Main changes in Version 1.7.1 (2018-07-07):

  • faster implementation

Main changes in Version 1.6.5 (2017-12-24) --- this is only a minor release:

  • bug fix in the log2-cpm transformation computation (for preprocessed = FALSE) in sp_weights
  • bug fix in sp_weights when facing NA or NaNvalues

Main changes in Version 1.6.3 (2017-11-09)

  • WARNING instead of ERROR when a gene set with no measured genes is tested
  • NA support through na.rm_... logical arguments
  • bug fix for row.names error in weights computation for gene-wise analysis

Main changes in Version 1.6.2 (2017-08-04)

  • bug fix to the log2-cpm transformation computation (for preprocessed = FALSE)
  • improvement to defaults in varseq and tcgsa_seq wrapper functions, including an option to compute heteroscedasticity weights without conditioning on the variable(s) to be tested

Main changes in Version 1.5.2 (2017-07-11)

  • user-friendly wrapper function varseq for gene-wise testing
  • RAM usage optimization

Main changes in Version 1.5.1 (2017-05-24) --- this is only a minor release:

  • homogeneous test now available for gene-wise testing.

Main changes in Version 1.5.0 (2017-03-27):

  • permutation test now available for gene-wise testing

Main changes in Version 1.4.0 (2017-02-21):

  • bug fix for gene-wise testing

Main changes in Version 1.3.1 (2016-11-24):

  • bug fix for testing several covariates together (such as several bases of time) which now gives correct p-values
  • support for gene-wise testing
  • example dataset 'baduel_small' was changed to 'baduel_5gs' which includes 1943 additional gene expressions and the definition of 5 gene sets
  • documentation updated

Main changes in Version 1.2.0 (2016-07-19):

  • support homogeneous gene set test
  • documentation updated

Main changes in Version 1.1.0 (2016-05-26):

  • improved imports
  • documentation updated

Main changes in Version 1.0.2 (2016-05-18) --- this is only a minor release:

  • documentation and CITATION updated

Main changes in Version 1.0.1 (2016-05-14) --- this is only a minor release:

  • help file improvements