Release v1.0.2

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

[1.0.2] - 2023-07-14

Fixed

• crash with combination of technical reps and biological reps when combining them.
• idr bug with numpy dependency > 1.20
• replacing all spaces with underscores in the samples.tsv
  • should only affect columns where this was not enforced already (custom columns)
  • required for rule multiqc_samplesconfig

Release v1.0.0

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

[1.0.0] - 2023-05-31

Added

• CRAM support for ATAC+ChIP+RNA-seq workflows (in addition to the existing alignment workflow cram support)

Changed

• sctk yaml simplified

Fixed

• DESeq2 should no longer crash without DE genes
• bug with single-ended reads and subread
• gimme maelstrom dependency missing
• gimme maelstrom bug when XDG_CACHE_DIR is not set

Release v0.9.9

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

[0.9.9] - 2023-04-21

Changed

• moved downloading fastqs to localrules
• bam indexes are kept (not automatically removed)
• Salmon updated to the latest version v1.10.1 (fixes a bug)
• upsetplot updated to the latest version (fixes a bug)
• genomepy updated to the latest version (no reason)
• tabulate updated to the latest version (longer python support)
• everything else updated to the latest version
• --snakemakeOption debug_dag=True can now be used with 1 core (required)
• creating conda environments now faster
  • updated conda & mamba
  • dropped indexing of Conda's defaults channel
• one less global variable! (sanitized_samples)
• dropped correlation scores from DESeq2 clusterplots
  • pheatmap is too finickey to get the fontsize right
• pheatmap uses the sample order (from the samples.tsv) as best as possible

Fixed

• edge case when a GSM sample is a reanalysis of another GSM sample.
• error message referring to --config while it should be --configfile
• cyclic dependency on rule samtools_sort (caused by tildes in config paths)
• bug in DESeq2 related rules when using custom assemblies
• clear error message when downloading single-end data annotated as paired-end.
• "Max retries exceeded with url" for CRX samples
• upsetplot & assembly_stats segfault due to interactive matplotlib backend
• DESeq2 error: "EOF within quoted string"
• conda environment channel priorities
• trackhub index generation now removes rogue spaces from the annotation (instead of crashing)

Release v0.9.8

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

Added

• (experimental) support for the Chinese Genome Sequence Archive (GSA). Samples can start with their CRX identifiers.
• a force_assembly_hub flag to make an ucsc assembly hub even though a trackhub already exists

Changed

• MultiQC version updated (1.14)

Fixed

• edge case with the downloading-fastq workflow when samples file has an assembly column
• workflow explanation not being properly added to qc report
• download-fastq finishing succesfully with an error message..

v0.9.7

[0.9.7] - 2023-01-03

Added

• a message at the end of a succesful seq2science run where to find the report and the trackhub
• nicer error when genomepy has trouble querying the providers
• flag infer_motif2factors for whether or not motif2factors should be run

Changed

• Snakemake backend updated to most recent version
• for atac-seq workflow macs2_keep_mates is enabled by default.
• Workflow DAGs in the documentation are now simplified
• MultiQC version updated
• Fastqs downloaded by seq2science are now removed when not used anymore, can be turned on/off with keep_downloaded_fastq
• updated gimmemotifs

Fixed

• fixed macos errors thanks to reporting of @Jerolen and @WouterVGKULEUVEN
• clear error when specifying unavailable trimmer (#888)
• fixed bug with rule combine_biological_reps when no biological reps/descriptive names are present
• issue computeMatrix_gene without configurable distances, visualized in multiQC plotProfile (#905; default now 3000 bp up&down stream of gene)

Release v0.9.6

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

Changed

• all conda environments now work with strict channel priorities
• singlecellTK environment updated (no longer needs pip)
• increased expected RAM usage of scRNA-seq rule sctk_qc

Fixed

• outdated dependency in scRNA-seq rule export_sce_obj
• error in singlecellTK script with negative count values in reportCellQC()
• softmask_track_1 should no longer hang indefinetly
• On UCSC assembly hubs, the softmask track should align better (fixed off-by-1)
• upsetplot environment being broken (matplotlib version pinned)
• deeptools environment being broken (matplotlib version pinned)

Release v0.9.5

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

Changed

• no longer writes multiqc filenames to an intermediate file
• Updated kb-python to 0.27.3

Fixed

• downloading fastq from ena directly fixed

Release v0.9.4

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

Fixed

• (hotfix) pinned the snakemake backend for working rerunning

Release v0.9.3

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

Added

• seq2science specific lockexception and cleanup metadata errors
• deseq2science now accepts the optional argument --assembly, which can be used if the samples.tsv contains >1 assembly to specify which one is used.
  • By default, the first assembly is used (same as before)

Changed

• rules that download something get re-tried once, in case internet is unstable
• bam files are no longer copied when sieving is not required
• moved blacklist rules to blacklist.smk
• rule inputs now use rules.rulename.output where possible
• renamed .smk files to match the naming schemes of the other .smks.
• added additional comments to clarify what happens to bam files
• cleanup cache+tarballs of conda environments, saving lots of precious disk space

Fixed

• fixed custom assembly extensions (e.g. ERCC spike-ins) for scATAC-seq and scRNA-seq
• profiles work again
• deseq2science now has a clear separation between positional and optional arguments
• issue with blacklist bed containing more than 3 columns

Release v0.9.2

Automated preprocessing of Next-Generation Sequencing data, including full (sc)ATAC-seq, ChIP-seq, and (sc)RNA-seq workflows.

Added

• seq2science specific lockexception and cleanup metadata errors
• deseq2science now accepts the optional argument --assembly, which can be used if the samples.tsv contains >1 assembly to specify which one is used.
  • By default, the first assembly is used (same as before)

Changed

• rules that download something get re-tried once, in case internet is unstable
• bam files are no longer copied when sieving is not required
• moved blacklist rules to blacklist.smk
• rule inputs now use rules.rulename.output where possible
• renamed .smk files to match the naming schemes of the other .smks.
• added additional comments to clarify what happens to bam files
• cleanup cache+tarballs of conda environments, saving lots of precious disk space

Fixed

• fixed custom assembly extensions (e.g. ERCC spike-ins) for scATAC-seq and scRNA-seq
• profiles work again
• deseq2science now has a clear separation between positional and optional arguments
• issue with blacklist bed containing more than 3 columns