Memote

------------------
* Distribute the missing tests.

------------------
* Fix a problem with the report caused by previous refactoring.

------------------

* Remove the pip dependency in ``show_versions``.
* Update the CI to use stages and ``tox-travis``.
* Modify some editor and other configuration.
* Expose testing of experimental essentiality and growth data in memote.
* Create a configuration system for media that is extensible to further
experimental data types.
* Add test for identifying purely metabolic reactions with fixed constraints in
models
* Add test for identifying transport reactions with fixed constraints in models
* Add test for identifying reversible oxygen-containing reactions in models
* Add division import from __future__ to ``test_biomass`` and
``test_consistency`` where it was missing.
* Add O2 to MetaNetX shortlist, allowing for easier identification
* Allow tests and test module to be skipped or run exclusively when creating
a snapshot report.
* Fix some typos
* Add history report view and connect it to `memote report history` call.
* ``find_direct_metabolites`` detects and removes false positives.
* ``find_transport_reactions`` detects reactions using forumlae and annotations
* Add tests for detecting gene annotations (and verifying they are in
MIRIAM style)
* Add unit tests for ``matrix.py`` in file ``test_for_matrix.py``.
* Add tests ``find_metabolites_not_produced_with_open_bounds`` and
``find_metabolites_not_consumed_with_open_bounds``
* Add test ``find_duplicate_metabolites_in_compartments`` to detect duplicate
metabolites in identical compartments
* Cache heavily used support functions in ``helpers.py`` and
``consistency_helpers.py``

------------------

* Test summary only displays extended narrative summary describing test,
and not one-line summary describing expected function behavior/output
* Fix the following bugs:
- Fix type annotation on the test for Biomass Production in Complete Medium
- Fix TypeError when running memote new which was associated with unicode
and string formatting in py2.7
- Sort existing test results from misc into the respective categories
(by editing test_config.yml)
- Move Matrix statistics category to unscored side into their own card
- Add a tuple of (number of reactions, number of genes) to the data
annotation of the metabolic coverage test.
* Add filter in ``report_data_service`` that changed the test result status to
"error" when the data attribute is ``null``, thus avoiding that the report
interface breaks when trying to access data.
* Add test for identifying stoichiometrically balanced cycles in models
* Correct the arguments used for repositories such that ``memote run`` and
``memote history`` work as expected inside of a repository.

------------------

* Emergency fix for distributing required JSON file.

------------------

* Let Travis re-package the snapshot report with every release.
* Add new module to test for the presence of SBO term annotations.
* Add a test for Biomass production in complete medium.
* Clarify extend of mass- and charge-imbalance testing.
* Remove much of the boilerplate code of the report template as a preparation
for the history and diff report.
* Fix bug with test_blocked_reactions
* Update the testData.json with data from the previous release
* Fix a small bug with the metrics of mass/charge unbalanced reactions.
* Correctly invert the found identifiers in wrong annotations and namespace
consistency in order to report the correct results.
* Add a cross-reference shortlist using MetaNetX flatfiles
* Add a script that can be used to add more metabolites and then to
re-generate the shortlist
* Add helper function ``find_met_in_model`` which looks up a query metabolite
ID using the MNX namespace in the shortlist and:

- If no compartment is provided, returns a list of all possible candidates
metabolites.
- If a compartment is provided, tries to return a list containing only
ONE corresponding metabolite.

* Add helper function ``find_compartment_id_in_model`` to identify
compartments using an internal shortlist of possible compartment names.
* Provide tests for each function
* Refactor code to use these functions specifically:
- ``find_ngam``
- ``find_biomass_reaction``
- ``detect_energy_generating_cycles``
- ``find_exchange_rxns``
- ``find_demand_rxns``
- ``find_sink_rxns``
- ``gam_in_biomass``
- ``find_biomass_precursors``
* Improve ``find_ngam`` in addition to agnostically looking for ATP hydrolysis
reactions, the test now also looks for a range of possible "buzzwords" in
the reaction NAME: ['maintenance', 'atpm', 'requirement', 'ngam',
'non-growth', 'associated']. One match suffices as a classification.
* Improve ``find_biomass_reaction`` to look for three attributes in a biomass
reaction, one of which is sufficient to classify it as a biomass reaction:

1. "Buzzwords" in the reaction ID: ['biomass', 'growth', 'bof']
2. An annotation matching the SBO-Term SBO:0000630 specifically!
3. Containing a metabolite matching the regex:
``^biomass(_[a-zA-Z]+?)*?$`` (case-insensitive)
* Add function ``bundle_biomass_components`` to identify whether a given
biomass reaction is 'split' or 'lumped'. This function looks simply at the
size of the biomass reaction. Based on a guess-timated cut-off the reaction
is then classified. If it is 'lumped' it is returned without changes, if it
is 'split' the reactions of any non-energy precursor metabolite are returned
as well. This is based on the assumption that a 'split' biomass reaction has
the following structure:
a (1 gDW ash) + b (1 gDW phospholipids) + c (free fatty acids) +
d (1 gDW carbs) + e (1 gDW protein) + f (1 gDW RNA) + g (1 gDW DNA) +
h (vitamins/cofactors)-> 1 gDCW biomass.
We're supposing that for each macromolecule precursor metabolite there is a
single reaction defining its composition i.e. ``e`` = protein would have the
reaction: ``alanine + asparagine + ... + valine --> e``
* Add function, test and model test to identify missing essential precursors
to the biomass reaction.
The function is ``essential_precursors_not_in_biomass``
* Record the score of individual test cases and sections in the result output.
* Correct the import of module 'annotation' with 'sbo' in ``test_sbo.py``
* Refactor sink_react_list to sink_reactions for improved readability
* Allow ``test_sink_specific_sbo_presence`` to be skipped when no sink reactions
are present with a metric of 1.0
* Fix a bug that compared the length of a float to generate a metric in
``test_basic.py`` and generated a TypeError.
* Fix a bug that prevented ``find_biomass_precursors``
in ``memote/support/biomass.py`` from functioning due to a malformed set
* In CONTRIBUTING.rst replace link to semantic commit guide by seesparkbox
with link to guide by karma, due to error with sphinx linkcheck.
* Fix a bug that prevented ``find_biomass_precursors`` from correctly
identifying ``atp`` and ``h2o`` metabolites in cobra model reactions
* Fix improperly labeled sbo terms for biomass production in ``biomass.py``
and ``test_for_helpers.py``.
* Add matrix conditioning functions in ``matrix.py`` which are used for
model stoichiometric matrix testing in ``test_matrix.py``
* Add missing rank and nullspace_basis functions in ``consistency_helpers.py``
* Fix issue with improper string/dict formatting in ``test_biomass.py`` tests
* Re-organize the architecture to read in external configurations and add
custom tests.
* Add an argument ``--location`` which replaces ``--directory`` which can be
used to set the directory or database where results should be stored.