infercnv release v1.3.3

Significant speed up of plot_cnv(), more significant the bigger the matrix. Now also reuses stored hclust for references.

InferCNV Release v0.99.7

Numerous fixes have been made.
Improved when subclustering is done to be more accurate.

InferCNV Release v0.99.0

This is a large update to inferCVN, including:
-more flexible infercnv::run() method
-resume-level functionality, so it will reuse existing processed data objects on re-running with different parameters.
-additional denoising methods included
-CNV predictions using hidden Markov models.

The wiki documentation has been heavily revised to reflect the updated functionality and usage.

InferCNV Release v0.8.2

uses zero-inflated negative binomial to simulate spike-ins used for scaling.

InferCNV Release v0.8.1

patch release - fixes a bug that impacted the multi-patient view where some cells were switching between patient panels, dependent on a mixed ordering within the cell annotations file.

InferCNV Release v0.8.0

This is a major update to InferCNV. See updated wiki documentation for new usage info.

InferCNV Release v0.3

-includes additional options: 
    
       --ref_subtract_method=REFERENCE_SUBTRACTION_METHOD
		Method used to subtract the reference values from the observations. Valid choices are:  by_mean, by_quantiles  [Default by_mean]
      
     	--hclust_method=HIERARCHICAL_CLUSTERING_METHOD
		Method used for hierarchical clustering of cells. Valid choices are:  ward.D, ward.D2, single, complete, average, mcquitty, median, centroid  [Default complete]


The --steps parameter now generates a full inferCNV heatmap/plot for each of the data transformation operations performed.

The log transformation log2(x/10 + 1) to generate transcripts(or counts) per 100k instead of per million is now more simply  log2(x+1).  If the user wants to study counts-per-100k or counts-per-10k, that is entirely fine...   The log transformation will simply be log2(whatever + 1).

-sample data is updated using a random selection of 400 malignant oligodendroglioma cells and ~100 normal cells, described in file '__sampled_cells.annotations.dat'.  The actual gene names and cell names are provided instead of the generic gene_ and cell_ values.  Data is from Tirosh et al. Nature 2016.

infercnv-v0.2

Now, works with R version >= 3.2

InferCNV as a library or script.

InferCNV can now be installed as a library from the associated tar.gz or directly from GitHub.

If installing from the tar.gz, use the following command on command line.

R CMD install infercnv_0.1.tar.gz

If installing from GitHub, use the following command in R.

library("devtools")
install_github("broadinstitute/inferCNV")