Release 2.5
Changes in Cpipe 2.5.0
Command Line Interface Rework
cpipe commands have been completely rewritten in Python to have much more functionality and to be much more user friendly. Features have been added to existing commands, like the ability to create a batch using cpipe batch create --data *.fastq, and the ability to run the pipeline with bpipe options , using cpipe run MyBatch -n 3
New commands include:
cpipe stop, which stops a currently running batchcpipe batch check, which validates a batch metadata file against a comprehensive schemacpipe batch editandcpipe batch view, which start an interactive editor for editing/viewing the sometimes complicated batch metadata files.
For more information, refer to the command documentation.
Environment Shell
It is also now possible to start a bash shell with its environment set in such a way that it is able to use Cpipe libraries and binaries on its $PATH, including tools such as samtools, bedtools, vep, bpipe etc. This can be useful for debugging Cpipe runs, for running extra Cpipe utility functions, or even for using Cpipe tools for other purposes. For more information, refer to the documentation.
New VEP Script
Ensembl VEP, which is used by Cpipe for variant annotation, has been updated to the new reworked version, which should be much faster and more accurate. The VEP cache has also been pre-indexed, which should also add to speed improvements.
Public Install
It is now possible to run a full automated install, without needing the MGHA credentials. This means that using Cpipe is again very easy for any member of the public. All you need to do manually is find a copy of GATK.
Docker Registry
MGHA users can now install Cpipe using Docker, by pulling from our Docker registry. Refer to the Docker registry documentation here for more information.
Interactive Bpipe Config Generation
The installation script will now prompt you for your queueing system, and give you some configuration options, which should make installation much simpler and less error prone.
Miscellaneous Changes
- Pipeline IDs are now automatically generated
- Large documentation changes
- Fix 1 less sample being reported by the pipeline than was actually found
- Support for nested batches (a batch inside a directory inside
batches) - Fix potential infinite loop if the user manually
sourced theenvironment.shfile - Fixed issues with the Torque executor
- Python code was restructured into a
distutilspackage, meaning that code is now properly shared between scripts, allowing more maintainable code - Use updated metadata format, which has two optional extra columns
- Updated tools:
- Python to Python 3.6
- Bpipe to 0.9.9.4
- Vep plugins updated
- Pip packages updated
- Various other bug fixes