nf-core/scnanoseq

Single-cell/nuclei pipeline for data derived from Oxford Nanopore and 10X Genomics

10xgenomicslong-read-sequencingnanoporerna-seqrnaseqscrna-seqsingle-cell

Version history

Special thanks to a new contributor to scnanoseq:

#94 Strict syntax conversion: converted entire workflow to strict syntax and reorganized into <name>/main.nf directory structure
#93 and #55 Upgraded IsoQuant from v3.6.1 to v3.13.0 and removed chromosome-splitting logic in the IsoQuant subworkflow due to improvements in IsoQuant; IsoQuant now processes all chromosomes in a single invocation
#65 and #61 Replaced NanoFilt with Chopper for read filtering, with gzip compression of intermediate split FASTQ files to reduce disk usage
#87 Added --skip_blaze_demux parameter to allow skipping BLAZE demultiplexing
Added SPLIT_SEQ module using seqkit for splitting FASTQ files, replacing the previous split approach
Updated CAT_FASTQ to the nf-core module which now supports compressed inputs
Moved gzip compression to pre-extraction step to minimize uncompressed FASTQ footprint in work directories
Upgraded nf-core template from 3.2.1 to 3.5.1

--skip_fastqc default changed from false to true (disabled by default due to runtime issues with long-read data)
--skip_fastq_nanocomp default changed from false to true
--skip_bam_nanocomp default changed from false to true
#99 --skip_toulligqc default changed from false to true
--skip_blaze_demux added (default: true)

#83 Isoquant count matrix headers are now compatible with Seurat (removed leading # from header line). Note that lastest version of IsoQuant also resolves this issue.

Fixed error that would occur when a large amount of transcripts were used to split a bam by replacing samtools_view with a custom script.

(#76) Fixed issue where -resume would not always cache IsoQuant steps, resulting in silently skipping chromosomes

(https://github.com/nf-core/scnanoseq/issues/44) All output files produced by isoquant are now produced in the results file
(https://github.com/nf-core/scnanoseq/issues/47) BLAZE scripts has been removed from the repo so the actual published code can be used
(https://github.com/nf-core/scnanoseq/issues/47) Added new whitelists for 10X 3v4 and 10X 5v3
Upgraded nf-core template to 3.2.1
Upgraded nanocomp nf-core module (no version change)
Seurat process now places the seurat object to pipeline outputs
Updated metro diagram

(https://github.com/nf-core/scnanoseq/issues/45) Reference files are now accepted in .zip format
(https://github.com/nf-core/scnanoseq/issues/56) Fixed an error where using --skip_dedup would end the pipeline early
(https://github.com/nf-core/scnanoseq/issues/58) Fixed UMI length for 5 prime chemistries
Fixed an error caused by --skip_qc and --skip_seurat
No longer output uncorrected correct barcodes
Fixed issue with linting.yml preventing automatic template PRs

Dependency	Old version	New version
`BLAZE`	2.2.0	2.5.1

Inputs for IsoQuant are split on chromosome to allow for faster processing
The read counts QC metric is now able to leverage NanoPlot counts if FastQC is skipped
Added oarfish as an option for quantification
Added picard markdupes as an option for deduplication

The ‘Post Trim Read QC’ and ‘Post Extract Read QC’ nodes on the metro diagram have been placed in correct locations (closes issue #36)
The BLAZE process in the example config has been corrected to use cpus instead of --threads