- Remove PCR chimerism in single-cell data: for each cell barcode - UMI combination, only keep lineage barcode with the most reads; remove all ambiguous ties
- Replace fastx_toolkit with fastp for filtering reads based on quality
- Add option for filtering reads: fastp low complexity threshold
- Filter R2 (read containing lineage barcode) in single-cell data with fastp if input data is fastq files. By default, R2 will be filtered for >= 80% bases with >= phred score 20.
- Update conda environment and Docker image with fastp dependency (v1.4). Conda environment must be updated and Docker image must be pulled as described in
README.md. - Run starcode_umi, not starcode, when clustering unmapped reads from single-cell data
- Update and improve documentation
- Output location for clustered unmapped reads was moved to
counts/unmapped/ - Publishing mode for all files that are copied to the results folder was set to
overwrite: true.
-
Clustering unmapped reads
- always save unmapped reads to fastq file with bowtie
- lower resource usage for starcode on unmapped reads
- trim reads before running starcode on unmapped reads (if only upstream or only downstream adapter trimmed)
-
add NEWS.md to track changes
-
fix resource labels in slurm and lsf config