Release 4.0.3

Changelog

• Fixed issue with BAM processing
• Create output directory if it doesn't exist
• Tuned disk requirements for several jobs
• Fixed toil-rnaseq-inputs compression of workflow inputs
• Fixed rsem_gene_mapping to support zero mapped genes for non-human models

Release 4.0.0

Major Changes

• BAM support
• GDC support: BAM IDs can be supplied via the gdc:// URL
• New tool added: Hera
• Added option to generate workflow inputs derived from users reference genome and annotation files: toil-rnaseq-inputs

Minor Changes

• Added user-defined config option: toil-rnaseq config-input
• Added user-defined manifest option: toil-rnaseq manifest-input
• Added documentation for Auto-scaling on AWS
• Colocated all external dependency code within the repository (except Toil)
• Entire workflow DAG rewritten statically for clarity
• Default workflow inputs supplied in config now open-access
• Switched PyPi pushes to Twine
• Removed BamQC
• Documentation overhaul

Release 3.3.3

• Remove suffix from output of single-end samples (#96)
• Misc. fixes to production wrapper
• Cut down on redundant sdist builds
• Push short tag for production Docker container

Release 3.3.0

New minor release that addresses memory issues when sorting with STAR. Pipeline is faster if bam isn't being saved or if it is being saved in conjunction with BamQC.

• Use samtools to sort if bam is being saved (#84)

Release 3.2.5

• Fixed bug when running in single-end mode (#72)
• Ensured restart capability is implemented (#76)
• Improved robustness of output-dir handling
• Fixed None bug when deleting FileStoreIDs (#80)

Release 3.2.2

• Fusion support added from Kallisto — kallisto subdir now contains a fusion.txt file as part of standard output which can be analyzed with their tool: pizzly.
• Intermediate files generated by the pipeline are now aggressively deleted, which should drastically reduce the size of the jobStore during runtime, especially with many samples.
• Fixed bug where a variable was initialized to None instead of an empty list when files were submitted as fastqs.

Release 3.2.0

• Reverted a dependency that caused discontinuity of expression results between V2 and V3 of the pipeline
• Splice Junction (SJ.out.tab) from STAR now part of standard output
• Now relies on toil-lib==1.1.5
• Updated README to reflect output changes

Release 3.1.0

A new release of the toil-rnaseq pipeline, mostly focused around improvements for QC

• Added a subdirectory for STAR which contains log.final.out
• Added the file used for flagging a sample as pass/fail from the bamQC output
• Added a processing speedup if the sample is submitted as an unzipped fastq pair
• Documentation updates, including a tree diagram of the complete toil-rnaseq pipeline output
• Relaxed heuristic for fastq naming convention when determining paired files
• Added the toil-rnaseq CWL tool descriptor
• Moved the Dockerfile and wrapper code to this repository
• Added disk requirements for all jobs to avoid caching errors

3.0.2-3

Fix gzip bug

3.0.2-2

Fix to allow a JSON input with no tar file input