pwlb/drop-seq-pipe repository overview

This is a docker image based on the dropSeqPipe⁠. The image can be used to process dropSeq or seq-well data sequenced with Illumina (and possibly other formats). There is some additional functionality including merging fastq.gz files and automatically creating the samples.csv file required by the dropSeqPipe by extracting them from the names of the input files.

⁠Usage

⁠Environmental variables

Under normal circumstances no

The environmental variables used in this container include

• SAMPLENAMES: the names of the samples, that will be used to select and merge fastq.gz files. For example

  SAMPLENAMES=sample1 sample2
  

  Copy

  would merge all the fastq files starting with sample1 and sample2. By default SAMPLENAMES="" (i.e. if not specified) and sample names will be determined automatically by extracting the sample names with the regular expression r"(.*)_S\d+_L\d{3}_R\d_\d{3}.fastq.gz".

• NUMCELLS: a number of cells/beads to extract from each sample (counted after merging the fastq files). If you don't want to discard any barcodes set this to a large number but this may raise some errors at the last stages of dropSeqPipe where it tries to merge all count tables into one (but you can safely ignore those and just use the intermediate count tables).

• JOBS: number of processes to be used by snakemake.

• TARGETS: the type of analysis to perform, the default is all. If you want to perform just a preliminary qc select one of the available targets from https://github.com/Hoohm/dropSeqPipe/wiki/Running-dropSeqPipe#modes⁠.

⁠Volumes

The pipeline expects the following volumes to be mounted

• /input:ro, the location of the fastq.gz files
• /output:rw, where the results will be stored
• /meta:rw, the location of the annotation and genome files. The script will look for transctripts.fasta and annotation.gtf files. Then it will generated STAR index and other files necessary for the pipeline, if such files already exist the pipeline will reuse them without regenerating them.
• /samples.csv (optional), if you want to provide a specific samples.csv file according to the dropSeqPipe standard. If not present it will be automatically generated based on SAMPLENAMES or on file names if SAMPLENAMES is not provided.