Monocle 2.0.0 Series NEWS
================================================================================
Version 2.6.1
--------------------------------------------------------------------------------
  o knn-based density peak clustering is not general for all datasets. Rolled back to 
    the previous densityPeak clustering algorithm and set it to be the default algorithm.
    A new Louvain clustering algorithm for dealing with large datasets (> 50 k cells)
    is added. 
  o A few bug fixes for importCDS, exportCDS functions.


Version 2.6.0
--------------------------------------------------------------------------------
  o Official release of the features in 2.5.0 through BioC


Version 2.5.0
--------------------------------------------------------------------------------
  o Various plot utilities for visualizing complex developmental trajectory 
    (plot_complex_cell_trajectory), multi-way kinetic curve and multi-way heatmap
    (plot_multiple_branches_pseudotime, plot_multiple_branches_heatmap). 
  o ClusterCells now in principle supports clustering for 100 k+ cells which depends on 
    a procedure of kNN based density peak clustering algorithm. densityClust package 
    version 0.3 are required for users to use this functionality.  
  o New functions supporting importing scater or Seurat cds into monocle or vice versa
    (importCDS, exportCDS).


Version 2.4.0
--------------------------------------------------------------------------------
  o The default expressionFamily is now negbinomial.size, instead of Tobit. If 
    you are using TPM or FPKM data, we urge you to convert it to relative 
    transcript counts with relative2abs and use the negative binomial distribution
    in your CellDataSet objects.
  o Revamped clusterCells functionality based on t-SNE and densityPeak
  o New procedure for selected ordering genes called "dpFeature". See vignette
    for details.

Version 2.1.1
--------------------------------------------------------------------------------
  
BUGFIXES
  o Fixed an problem where reduceDimension would return different results on 
    repeated runs given the same inputs.The problem was actually in DDRTree in 
    two places: kmeans and irlba. We now call irlba with deterministically 
    initialized eigenvectors and kmeans with deterministically selected rows of 
    the input.
    
  o Fixed a problem in classifyCells related to joining factors and levels. This
    would generate annoying warnings.
    
  o Fixed the check for valid sizeFactors prior to differentialGeneTest. Without
    this check, differentialGeneTest would report FAIL on all genes because of
    factors not having enough levels. 

Version 2.1.0
--------------------------------------------------------------------------------
FEATURES
  o Monocle's algorithm for converting relative expression values (e.g. TPMs)
    into absolute transcript counts, called "Census" has been re-designed. The
    new version is much more accurate. Census is available through the 
    relative2abs() function. The interface to this function has changed, and
    the output values will be quite different from the old version. The main
    change is that the version in 1.99.0 reported cDNA counts, as estimated 
    in each cell's cDNA library. Now, relative2abs reports estimates of mRNA
    counts in the lysate. These are easier to compare with experiments that
    estimate mRNA counts via the use of spike-in controls.
    
  o A new heatmap function plot_pseudotime_heatmap() replaces the old 
    plot_genes_heatmap().
    
  o Lots of new documentation. More to follow in upcoming releases.
    
BUGFIXES
  o Ordering cells with DDRTree no longer compresses cells at the tips of
    trajectories.
    
  o Pseudo counts are now applied correctly no matter what the underlying 
    distribution used to model expression is.
    
  o Variance stabilization is applied correctly when a CellDataSet object's
    expressionFamily is negbinomial.size().
    
  o The gene_id field in results from calculateMarkerSpecificity is now a
    character instead of a factor. This was leading to indexing errors and 
    nonsensical semi-supervised clustering and ordering results. 
    
  o BEAM() and plot_branched_heatmap() now use cBind instead of cbind, fixing an 
    issue with sparse matrices.
    
Version 1.99.0
--------------------------------------------------------------------------------
  o The first public release of the Monocle 2 series. For a summary of 
    new features and changes, please see: 
    http://cole-trapnell-lab.github.io/monocle-release/features/


CHANGES in monocle 1.1.5:

* After grappling with Bioconductor build issues in HSMMSingleCell related to
  VGAM updates, we have moved to a different data layout in HSMMSingleCell, 
  which has caused some changes in the vignette.

CHANGES in monocle 1.1.1:

* Fixed a bug in responseMatrix() that occurs when you don't have any genes that fail VGAM fitting