Second beta bug fix

Fixes #7: genes in both target_a_id and target_b_id are now correctly identified in scoring notebook.

0.4.0:
This release primarily provides improved error handling and reporting. Additional user-facing changes are:

• An all-in-one installation script is now provided to simplify installation further.
• Software now correctly expands input relative paths, as well as those containing '~' or environment variables.
• Notebook default paths have been changed to be cross-platform compatible.
• A custom.css file is provided with the notebooks to format them for attractive printing.

Second beta release

This release primarily provides improved error handling and reporting. Additional user-facing changes are:

• An all-in-one installation script is now provided to simplify installation further.
• Software now correctly expands input relative paths, as well as those containing '~' or environment variables.
• Notebook default paths have been changed to be cross-platform compatible.
• A custom.css file is provided with the notebooks to format them for attractive printing.

First beta release

This release primarily addresses installation and packaging. The main change visible to the user is that the software can now be installed directly through conda as the package dual_crispr.

Other user-facing changes include:

• Set-up, count, and scoring commands can be called by simply running set_up_dual_crispr, count_dual_crispr, or score_dual_crispr; direct invocation of the script names through python is no longer necessary.
• Raw data, temporary interim data, and processed output are no longer required to be placed in subfolders of defined raw, interim, and processed directories; users now specify input and output locations as command line arguments. Temporary interim data is automatically placed in a separate folder at the output location.
• A bash script is now provided to automatically set up a Jupyter Notebook server through which the notebooks in the software pipeline can be accessed via a remote web browser. While browser-based access is not necessary for the software to function (it can be used solely from the command line), it is a convenience.
• Test data files have been considerably augmented, and the default values in each notebook now reference these test files, allowing users to open the notebooks and run them individually against test data to see how they work (if desired).

In addition to these user-facing changes, the former utilities submodule has been broken out into a separate and independent library project, and is now installed as part of the dependencies.

Second alpha release

This release primarily addresses usability issues raised by a tester of the first alpha release. New features:

1. All information necessary to define construct libraries is now read from text files that can be easily created by the user; adding new libraries to the software only requires adding new files to the designated library_definitions folder.

2. All user-tunable settings (including folder locations of data and software files, number of processors on which to run the software, scaffold sequences, etc) are now read in from a configuration file. A configuration file with sensible defaults is provided during the installation to support novice users, while experienced users can modify this file to control most parameters used in the code.

3. Test data is now provided during installation, and commands to run the pipeline on this test data are provided in the quickstart guide. This test data should allow users an easy way to trial the software.

4. The scoring preparation notebook has been extended in several ways:

   • More than one valid time point prefix can now be recognized, with the list of acceptable prefixes being defined in the config file. By default, the software will recognize t, T, d, and D.
   • The software no longer requires that the sample name section of a column header contain only alphanumeric characters.
   • Format errors detected during scoring preparation are now reported all at once, instead of each format problem causing a separate fatal error (and thus requiring multiple runs of the scoring preparation notebook to detect). This should allow users to fix all the formatting problems in their input files at one time.
   • An informative error is now provided if multiple data columns in the input appear to contain counts for the same time point and replicate.

5. Notebooks whose primary endpoint is the creation of new files now report the name of the created file as their last step, in order to make clear that the processing has finished successfully.

In addition to these user-facing extensions, unit tests coverage for the python modules has been greatly increased, to 115 tests.