JBrowse 1.16.0

Major improvements

• Added ability to view paired read data as connected entities for BAM and
  CRAM store classes. There are multiple different viewing options for this
  including plotting by insert size, plotting as connected arcs, or pileup
  views for the paired reads. Additonally multiple color schemes are available
  for coloring by insert size, pair orientation, mapping quality, and more.
  Thanks to @garrettjstevens, @rbuels, @AndyMenzies, and @keiranmraine for
  testing. Also a big thanks to @jrobinso from @igvteam for contributions to
  CRAM code related to paired reads (pull #1235, @cmdcolin)

Minor improvements

• For users with the "dev" or compiling JBrowse from source code, the ./setup.sh
  now performs a full webpack production build (pull #1223, @cmdcolin)

• Created new BAM parsing mechanism using the npm module @gmod/bam.
  Users might see some modest performance improvements due to enhanced tooling.
  Thanks to @rbuels and others for testing and feedback (pull #1215, issue #1178,
  @cmdcolin)

• setup.sh now supports setting a JBROWSE_PUBLIC_PATH environment variable for
  more flexibility in iframeless embedding scenarios (issue #1213, @rbuels)

• Added support for indexing arbitrary fields from GFF3Tabix files by setting
  nameAttributes in the track config e.g. nameAttributes=name,id,customfield.
  (issue #1115, pull #1222, @cmdcolin)

• Add support for generate-names to index VCF features that have multiple IDs in
  the ID column (@cmdcolin)

• Added documentation on embedding JBrowse in an iframe and in a div, including
  how to embed JBrowse using a custom JavaScript object as a configuration
  (pull #1228, pull #1243, @rbuels and @garrettjstevens)

• Added ability to render non-coding transcript types to the default Gene glyph
  which helps when a gene feature has a mix of coding and non-coding subfeatures
  (issue #1106, pull #1230, @cmdcolin)

• Created new VCF parsing mechanism using the NPM module
  @gmod/vcf. Thanks to @cmdcolin and
  others for testing and feedback (pull #1227, issue #1199, @garrettjstevens)

• Added ability to open "chrom.sizes" files from the Open sequence dialog
  (issue #1250, pull #1257, @cmdcolin)

• Added a config datasetLinkToParentIframe to make the dataset selector use
  window.parent for when jbrowse is in an iframe (pull #1248, @enuggetry)

• Improved error message that is displayed when a data file cannot be fetched
  via CORS (@rbuels)

• Added some word wrapping for long unbroken fields in the View details
  popups. Thanks to @luke-c-sargent for the idea (issue #1246).

• Added hideImproperPairs filter for Alignments2/SNPCoverage tracks which
  disambiguates from missing mate pairs (pull #1235, @cmdcolin)

• Added useTS coloring option for RNA-seq strandedness, similar to the useXS
  that existed previously (pull #1235, @cmdcolin)

• Added a --bgzip_fasta option for prepare-refseqs.pl and also the ability
  to index reference sequence names if they are manually specified as a
  fasta index e.g. refSeqs=genome.fai. Thanks to @FredericBGA for the report!
  (issue #1281, pull #1282, @cmdcolin).

Bug fixes

• Fixed issue where some generate-names setups would fail to index features.
  Thanks to @BioInfoSuite for reporting (issue #1275, pull #1283, @cmdcolin)

• Fixed issue with getting feature density from BAM files via the index stats
  estimation (issue #1233, @cmdcolin)

• Fixed issue where some feature mouseovers where not working properly (issue
  #1236, @cmdcolin)

• Fixed issue where instantiating JBrowse via standalone.js didn't work when
  in a production build with JBROWSE_PUBLIC_PATH overridden (issue #1239,
  @garrettjstevens)

• Small fix for issue where SNPCoverage would crash on some feature filters
  (issue #1241, @cmdcolin)

• Fixed issue where JBrowse Desktop was not able to access remote files (issue
  #1234, pull #1245, @cmdcolin)

• Fix issue where the Hierarchical track selector contained a bunch of blank
  whitespace. Thanks to @nathanhaigh for reporting! (issue #1240, pull #1253,
  @cmdcolin)

• Fixed issue where whitespace surrounding GFF3 attributes and attribute names
  was incorporated (issue #1221, pull #1254, @cmdcolin)

• Fixed issue with some GFF3Tabix tracks having some inconsistent layout of
  features (issue #1244, pull #1260, @cmdcolin)

• Fixed CRAM store not renaming reference sequences in the same way as other
  stores (pull #1277, @rbuels, @cmdcolin)

• Fixed bug where older browsers e.g. IE11 were not being properly supported via
  babel (issue #1259, pull #1267, @cmdcolin)

• Fixed bug where some files were not being fetched properly when changing
  refseqs. Thanks to @luke-c-sargent for the report (issue #1252)

• Fixed storeTimeout on CRAM files being unused which can result in excessive
  fetches (pull #1235, @cmdcolin)

• Fixed issue where JBrowse would load the wrong area of the refseq on startup
  resulting in bad layouts and excessive data fetches. Thanks to @hkmoon,
  @cmdcolin, and @garrettjstevens for debugging (issue #1190, pull #1235, pull
  #1187)

• Fixed issue where CRAM layout and mouseover would be glitchy due to ID
  collisions on features (issue #1271, @cmdcolin)

• Fixed parsing of certain bigBed files that were hanging on track startup
  (issue #1226, pull #1229, @cmdcolin)

• Fixed issue where some parts of a CRAM file would not be displayed in JBrowse
  due to a CRAM index parsing issue (@cmdcolin)

• Fixed an issue where BAM features were not lazily evaluating their tags
  (@cmdcolin)

Notable changes

• Rendering of features in popups, mouseover tooltips, and feature labels were
  made to escape HTML. If you are using literal HTML labels in these places
  then set the attribute unsafePopup, unsafeMouseover, or unsafeHTMLFeatures
  on your tracks. Thanks to @garrettjstevens (pull #1263, @cmdcolin).