Immunedb

* Arbitrary metadata can now be specified for samples. The only required
fields are `sample_name`, `subject`, and `study_name`.
* Identification now uses multiprocessing for each sample independently, rather
than one process per sample.
* Sequences with stop codons in the CDR3 can now be added to clones.
* V-length is now correctly calculated for 5' trimmed sequences.
* `immunedb_sql` now has the optional `--query` argument to run a query via the
CLI.
* Exporting via the CLI should now be faster.
* Insertions and deletions can now be exported for clones and sequences.
* Selection pressure is now provided with the sample analysis API call.

* Arbitary fields can be specified in metadata files. The only required fields
are now `file_name`, `study_name`, `sample_name`, and `subject`.
* Clones correctly filtered when functionality is not stipulated.
* Sequences that comprise trees can now be filtered by their copy number using
`--min-seq-copies`.
* Sequences with invalid bases will no longer cause identification to fail.
* Various local-alignment bug fixes.
* The number of instances in a clone is now stored in the database.
* Mutations are correctly exported when limiting to certain samples.
* Sequences can now be exported in AIRR compatible GenBank format with
`immunedb_export genbank`.
* Sequences can now be exported in Change-O format with `immunedb_export
changeo`.
* Genotyping with TIgGER can be run by passing the `--genotype` flag to
`immunedb_identify`. See the docs for more instructions.
* Selection pressure is now stored in a separate table for easier querying.
* Optional baseline regression tests added.
* Tree node features with no values are now excluded.

* Local alignment has been entirely rewritten to use bowtie2. This drastically
reduces the time necessary to locally align sequences.
* The `/samples/overlap/` API endpoint now properly returns clones when not
filtering by functionality.
* A new `--min-seq-copies` flag has been added to `immunedb_clone_trees` which
limits which sequences are included in trees based on their copy number.
* Re-creating trees of specifically specified clones now requires the `--force`
flag.
* Default arguments for all commands are now automatically populated.
* API calls to list subjects now provides unique sequences as well as copies
and instances.
* Arguments for `immunedb_clones` have been changed. Three methods are now
available, similarity based for B-cells, identical based for T-cells, and a
lineage-separation method.
* Instance counts are now stored for clones.

* Metadata is now specified in TSV files rather than JSON. A
`immunedb_metadata` command has been added to automatically create a template
file from FASTA/FASTQ files.
* Gene representation has been modified to be more flexible with non-standard
naming schemes.
* J-gene identification has been substantially changed. If anchoring does not
work, full-sequence matching is attempted.
* Clones can now be exported in the format expected by VDJtools with the
`immunedb_export vdjtools` command.

* Subclones are now properly assigned.
* The order of exported sequences is now consistent.
* Updates for consistency with Python 3 test cases.

* ImmuneDB can now process T-cell sequences.
* Clonal assignment now includes an optional "subclone" process for
locally-aligned sequences. Subclones are clones which share features of their
parent, but contain insertions or deletions. See the documentation for more
information.
* External clonal assignments can now be imported with `immunedb_clone_import`.
* Optional rollbar support has been added to `immunedb_rest` to track errors.
* Logging has been overhauled and is more consistent with best-practices.