CHANGES IN VERSION 1.4.4
-------------------------

BUG FIXES

    o Fix doc link in geo_characteristics() which affected the Windows build
        machines.

CHANGES IN VERSION 1.4.3
-------------------------

BUG FIXES

    o The Bioc 3.6 version of recount was missing an important commit.
        https://github.com/leekgroup/recount/commits/RELEASE_3_6 was missing
        https://github.com/leekgroup/recount/commit/a8a078a2389bb3da09a4a03796a231044f5f97b0
        which lead to https://github.com/leekgroup/recount-website/issues/10.


CHANGES IN VERSION 1.3.13
-------------------------

BUG FIXES

    o Changed reproduce_ranges() since disjoint exons are more useful than
        reduced exons for downstream analyses.


CHANGES IN VERSION 1.3.12
-------------------------

NEW FEATURES

    o Added the function read_counts().


CHANGES IN VERSION 1.3.9
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o Added citations for 
        https://www.biorxiv.org/content/early/2017/06/03/145656 and
        https://f1000research.com/articles/6-1558/v1 as well as mentions to
        them in the vignette.


CHANGES IN VERSION 1.3.7
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o add_predictions() was bumped to version 0.0.05


CHANGES IN VERSION 1.3.5
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o Vignette now uses the new BiocStyle::html_document that was recently
        released.


CHANGES IN VERSION 1.3.2
------------------------

NEW FEATURES

    o coverage_matrix() now has two new arguments: scale and round. Use
        scale = FALSE to get raw coverage counts, which you can then scale with
        scale_counts(). scale is set to TRUE by default, so the counts are
        scaled to a library size of 40 million reads. round is set to FALSE by
        default, but can be set to TRUE if you want to get integer counts, just
        as in the default of scale_counts().

CHANGES IN VERSION 1.3.1
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o Changed the default version argument of add_predictions() to 'latest'.
        Internally, that's still 0.0.03.


CHANGES IN VERSION 1.1.27
-------------------------

NEW FEATURES

    o Added the add_predictions() function which appends the predicted
        phenotypes to a RSE object downloaded with recount. The phenotypes
        were predicted by Shannon Ellis et al, 2017 (citation coming up soon!).

CHANGES IN VERSION 1.1.26
-------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o Changed the citation now that the recount2 paper has been published at
        http://www.nature.com/nbt/journal/v35/n4/full/nbt.3838.html.


CHANGES IN VERSION 1.1.25
-------------------------

NEW FEATURES

    o Added the function getRPKM() which can be used with
        RangedSummarizedExperiment objects from recount and from other sources.

CHANGES IN VERSION 1.1.24
-------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o recount_url now includes the URLs for the GTEx bigWig files.

CHANGES IN VERSION 1.1.19
-------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o coverage_matrix() now returns a RangedSummarizedExperiment object. This
        matches the behavior of recount.bwtool::coverage_matrix_bwtool() and
        is more consistent with the use of RSE objects in recount.

CHANGES IN VERSION 1.1.18
-------------------------

BUG FIXES

    o coverage_matrix()'s helper function .read_pheno() was failing for some
        projects.

CHANGES IN VERSION 1.1.16
-------------------------

BUG FIXES

    o Fixed a bug in the counts in coverage_matrix(). They were being 
        incorrectly multiplied by 100.


CHANGES IN VERSION 1.1.14
-------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o Completed the change to Gencode v25 annotation for exon and gene counts.


CHANGES IN VERSION 1.1.13
-------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o We dropped TxDb.Hsapiens.UCSC.hg38.knownGene completely from recount
         and will be using Gencode v25 instead.

CHANGES IN VERSION 1.1.12
-------------------------

BUG FIXES

    o Updated snaptron_query() to comply with recent changes in Snaptron.


CHANGES IN VERSION 1.1.8
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o Updated the package so you can now access TCGA data. Now there's over
        8 terabytes of data available in the recount project!

CHANGES IN VERSION 1.1.6
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o snaptron_query() can now access GTEx and TCGA data.

CHANGES IN VERSION 1.1.5
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o Snaptron changed from stingray.cs.jhu.edu:8090 to snaptron.cs.jhu.edu so
        snaptron_query() has been changed accordingly.


CHANGES IN VERSION 1.1.2
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o The function reproduce_ranges() now has the 'db' argument. By default
        it's set to TxDb.Hsapiens.UCSC.hg38.knownGene to reproduce the actual
        information used in recount. But it can also be used with
        EnsDb.Hsapiens.v79 to use the ENSEMBL annotation. Then with
        coverage_matrix() you can get the counts for either an updated
        TxDb.Hsapiens.UCSC.hg38.knownGene or for EnsDb.Hsapiens.v79 at the
        exon and/or gene levels as shown in the vignette.

CHANGES IN VERSION 1.1.1
------------------------

SIGNIFICANT USER-VISIBLE CHANGES

    o The vignette now describes how to download all the data, how to check
        exon-exon junctions by class, and how to use SciServer compute
        to access all the recount data (over 6 TB) via http://www.sciserver.org/


CHANGES IN VERSION 0.99.30
--------------------------

NEW FEATURES

    o Added the function snaptron_query() which queries Intropolis via Snaptron
        to find if an exon-exon junction is present in the data.


CHANGES IN VERSION 0.99.29
--------------------------

BUF FIXES

    o Fixed an bug in the vignette. Thanks to Michael Love for noticing it!


CHANGES IN VERSION 0.99.0
-------------------------

NEW FEATURES

    o Created the package skeleton for recount
    o Added the function reproduce_ranges() for re-creating the gene or exon
        level information used in the recount project.
    o Added the function abstract_search() for identifying SRA projects of
        interest by searching the abstracts.
    o Added the function browse_study() for opening a browser tab for further
        exploring a project.
    o Added the function download_study() for downloading the data from the
        recount project.
    o Added the function scale_counts() for properly scaling the counts before
        performing a differential expression analysis with the
        RangedSummarizedExperiment objects hosted in the recount project.
    o Added the function expressed_regions() for defining the expressed regions
        in a chromosome for a given SRA study.
    o Added the function coverage_matrix() for computing the coverage matrix
        based on the regions of interest for a given SRA study.
    o Added the function geo_info() for obtaining sample information from GEO.
    o Added the function find_geo() for finding the GEO accession id given a
        SRA run accession (id). This function will be useful for SRA projects
        that did not have GEO entries at the time recount's data was created.
    o Added the function geo_characteristics() for building a data.frame from
        geo_info()'s results for the characteristics.
    o Added the function all_metadata() which downloads all the phenotype data
        for all projects. This function can be useful for identifying projects
        and/or samples of interests.