Release v0.2.1

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

Added

• Chipseeker images in MultiQC report

Fixed

• Fixed issue with some samples not being findable/downloadable with s2s
• Fixed has_annotation always looking for annotation even if local files present
• Fixed bug where scatac-seq workflow was making fastqc reports per sample

Changed

• will try to UCSC gene annotations in Ensembl format (which uses gene IDs for the gene_id field, contrary to the UCSC format that uses transcript IDs. Wild huh?)

Release v0.2.0

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

Fixed

• Allow for same condition name across different assemblies & different controls

Added

• HISAT2 as aligner for RNA-seq
• splice-aware HISAT2 indexing for RNA-seq
• quantifier HTSeq for RNA-seq
• quantifier featurecounts for RNA-seq
• Salmon will output a gene-level TPM matrix as well
• added/expanded seq2science explain info (now covers RNA- and scATAC-seq too)
• sequencing strandedness may now be inferred automatically (unless specified in the config/samples.tsv)
• strandedness results are displayed in the multiQC under "Strandedness"
• a DEXSeq counts matrixs can now be generated with dexseq: True
• seq2science CLI now has the same reason flag as snakemake (-r/--reason flag)
• (re)added fnwi + rimls logos to the qc reports that went missing in seq2science migration

Changed

• rules and script names in RNA-seq. ex: txi.R is now quant_to_counts.R to better reflect its function
• quant_to_counts.R now converts salmon transcript abundances to gene counts identically to DESeq2
• STAR no longer outputs counts, and is no longer found under quantifiers
• gene counts are generated from (filtered) bams when using either STAR or HISAT2 as aligner and HTSeq or featureCounts are quantifier
• batch corrected gene counts are generated if a DESeq2 design contrast inclused a batch
• batch corrected TPM are generated if a DESeq2 design contrast inclused a batch, and quantification was performed using Salmon
  • for us in ANANSE, for instance
• seq2science explain now retrieves messages from explain.smk.
• seq2science explain now used profiles and snakemakeOptions.

Fixed

• the alignment workflow no longer uses strandedness
• seq2science CLI can now be run without cores with a dryrun or profile with cores
• Jenkins code style (now used mamba to install flake8)

v0.1.0 - 2020-07-15

Added

• bwa-mem2 as aligner
• new command-line option explain, which explains what has been done, and writes your material & methods section for you!

Changed

• change the workflow names, replaced _ by -. (download_fastq to download-fastq, chip_seq to chip-seq, atac_seq to atac-seq, scatac_seq to scatac-seq, and rna_seq to rna-seq)
• changed the way seq2science is called. Moved all the logic from bin/seq2science to seq2science/cli.py

Fixed

• Bug when merging replicates and having controls

v0.0.3 - 2020-07-01

Fixed

• bug when specifying 2 cores, which rounded down to zero cores for samtools sorting and crash
• edger environment was incompatible
• seq2science cache on sensible location + seq2science clean fixed

v0.0.2 - 2020-06-29

Fixed

• samtools using the correct nr of threads after update to v1.10

Changed

• The count table for ATAC/ChIP-seq peaks is now made from finding all peaks within a range of 200 bp, and taking the most significant one (gimmemotifs' combine_peaks) and extending the remaining peaks 200 bp. On this count table quantile normalisation, TMM, RLE and upperquartile normalisation with CPM is done. Downstream steps log transform these and mean center them. This however means that for broadpeaks no count_table is generated.
• Snakefmt -l 121 applied

v0.0.1 - 2020-06-17

Many minor bug- and quality of life fixes.

v0.0.0 - 2020-06-11

First release of seq2science!