Changes in version 2.0.0                        

  - New function in the workflow, optimizeMatacellNumber, which
    estimates the optimal value of the cellNum parameter passed to
    calculateP2G.
  - Using scrapper::clusterKmeans to implement deterministic algorithm
    of determination k-mean cluster. It is used to create metacells in
    the calculateP2G function.
  - Added to the output of calculateP2G function empirical p-values and
    FDR. The null distribution is calculated based on random links of
    peaks to the genes from other chromosomes.
  - Default ChIP-seq output from tfBinding is changed to version 2 which
    imposes more stringent cutoffs. We have increased the number of
    unique reads to 20M (previously 10M in version 1) and the number of
    peaks passing FDR < 1e-5 to 1000 peaks (previously 100 peaks in
    version 1). For ENCODE data, we now remove any samples with
    Audit.NOT_COMPLIANT or Audit.ERROR. Flags.
  - The argument method to calculateActivity has been deprecated.
  - Bug corrected in addLogFC. p-values were mistakenly transformed from
    log10 when it should be transformed from natural log when logFC_ref
    or/and logFC_condition was specified

                        Changes in version 1.5.1                        

  - checking for the duplicated gene names in the input gene expression
    SingleCellExperiment
  - validation of the regulon object passed to pruneRegulon, addWeights,
    and calculateActivity

                        Changes in version 1.0.1                        

  - the function addMotifScore returns correct values instead of NA.

                       Changes in version 0.99.20                       

  - the function aggregateAcrossCellsFast has been added.

                       Changes in version 0.99.11                       

  - new function aggregateAcrossCells has been added, which is
    implemented in c++ and makes run time for calculateP2G much shorter.

                       Changes in version 0.99.5                        

  - new function addLogFC which adds log fold changes of gene expression
    to regulons and significance statistics for differential gene
    expression
  - cellNum argument to calculateP2G set by default to 100 (previously
    it was 200).
  - in pruneRegulon the check to the uniqueness of gene names has been
    added.

                       Changes in version 0.99.0                        

Version number downgraded to 0.99.0 to meet Bioconductor requirements.

                       Changes in version 0.98.0                        

commit 12fabd3278b76711f4b2cbdcaca6794a918ffd83

  - transcription factor motif databeses used by addMotifScore function
    (human_pwms_v1, and mouse_pwms_v1) were replaced wih the new ones
    (human_pwms_v2, mouse_pwms_v2) which are up-to-date.

BUG FIX

  - clusters vector provided to the addWeights and pruneRegulon
    functions is checked for the presence of NA values. If they are
    present the function stops preventing segmentation fault and session
    abortion.

                       Changes in version 0.9.12                        

  - in calculateP2G staring position of the gene has been corrected.