TSS enrichment

Add TSS enrichment plot.

Calculate TSS enrichment per the ENCODE data standards, but with the
ability to specify the size of the region around TSS in which
enrichment is calculated.

Make SRR891268 the default peak/FLD reference.

Improve plot layout on iOS.

Change how plot help works.

If a line plot (TSS, peak plots) has no data because peak metrics or TSS
enrichment were not requested, omit it entirely.

Add deb/rpm support

0.7.0

Add deb/rpm support

More robust static builds, distribution

Get static compilation working on Debian (with a custom HTSlib build, at
least). Add distribution targets to Makefile. Fix samtools sort
invocation in example pipeline.

Make installation more flexible

Fix toggling of visibility of numeric sample IDs

Make Makefile more standard with DESTDIR and prefix

Move web template directory

Change fragment length distance table header

Add references to peak plots

You can now plot reference metrics on the peak plots. Like the fragment length distribution reference, you can choose to use the peaks we called on SRR891268, calculate an average from all the peak metrics you supply, or load a reference from a JSON, CSV, or TSV file.

Also fixed read group handling; if an aligned segment in the BAM file has a read group tag which is not represented in the file header, a new set of metrics will be started for that read group. At the end of processing, metrics for any read groups defined in the header that didn't have any reads will be dropped.

Improvements to documentation, examples

0.6.1

Update README, QC pipeline and example

ataqv

Rename project. Major UI update to the web viewer.

Renamed the project to ataqv, to avoid confusion with the ataqc tool
from the Kundaje lab.

Reorganized tabs.

Tables are now handled with DataTables, improving performance and
usability with a large number of experiments.

Plots are now made with D3.

Added plot showing how the fragment length distribution of each
experiment diverges from a reference. The reference distribution can be
supplied in a JSON file, calculated as the average of all the
experiments, or you can just use SRR891268, a sample from the original
ATAC-seq paper.

The fragment length distance of each experiment is also listed in the
quality indicators table.

Added the ability to smooth the fragment length distribution plot, and
switch its y axis scale between linear and exponential. The exponential
scale can reveal nucleosomal periodicity in the tails of the plots.

Plot highlighting is coordinated; mouse over a plot or legend and the
same experiment or sample will be highlighted in all plots.

Added the ability to specify the range of the fragment length
distributions when generating a web viewer instance.

Read groups in the alignment file can now be ignored.

Thanks to Peter Orchard for implementing the initial fragment length
distance calculation, and the hours of discussion, coding and testing
that led to the current version of the plot.

Add support for read groups in alignment files.

You can now get metrics for every read group in your alignment file. The
semantics of some arguments have changed a bit so that they either
override metadata inferred from read group tags, or provide defaults
when read groups are not used.

For example, --peak-file can now be set to auto to look for peak files
named after read groups, or be set to a single filename to use for all
alignments.

Please review the output of ataqc --help for details.

The handling of organisms and reference genomes has changed. Now our
default sets of autosomal references are based on organism, not a
particular reference, and contain both prefixed ("chr1") and
numeric ("1") chromosome reference names.