scNMTseq pipeline (RNA + DNA)

The Nextflow pipelines are under pipeline directory. The corresponding config files are under config directory.

Both scRNA and scDNA parts of the pipeline assume the following directory structure:

• data/
  • - raw/ (fastq, bam, and intermediatery cov files) - logs/ (logs generated during the pipeline) - metadata/ (all metadata, including rename.csv and sample.sheet.csv) - cooked/ (all data used for the post-processing, such as .NOMe.CpG.cov, and .complexity.csv) - results/ - reports/ (all reports from the pipeline, this is where MULTIQC will find its reports) - figures/ (all figures generated in the post-processing)
• pipelines/ (all .nf files)
• config/ (all config files used by the .nf files)

How to run

The DNA and the RNA pipelines can be run with

dataset=230731IMT;
nextflow -log data/${dataset}/logs/nextflow.log run ./pipelines/scnmtseq.dna.nf -c config/scnmtseq.dna.config -profile kcs -N onurcan.bektas@lmu.de --dataset_name ${dataset} -resume

dataset=230731ITR;
nextflow -log data/${dataset}/logs/nextflow.log run ./pipelines/scnmtseq.rna.nf -N onurcan.bektas@lmu.de --dataset_name ${dataset} -resume

The multiqc process in the DNA pipeline might not work. In that case you can run the MULTIQC manually with

 dataset=230731IMT;
 multiqc --outdir data/${dataset}/results/reports/ --filename ${dataset} --verbose --force data/${dataset}/results/reports/ -c config/multiqc_config.yaml  --interactive --sample-names data/${dataset}/metadata/renaming.csv

Before running the pipeline

1. Change the temp and cache directories inside the config/scnmtseq.dna.config (and other config files) according to your server's temp folders.
2. Change the resource settings (e.g. memory, cpus, etc.) for SLURM inside config/base.config according resources of your cluster
3. Create sample.sheet.csv and renaming.csv for the dataset on which you want to run the pipeline (see more on this below).

How to create the sample.sheet.csv and renaming.csv

Here are the typical contents of these files:

sample.sheet.csv is used to feed fastq files into the DNA pipeline, and renaming.csv file is used to rename the fastq files in MULTIQC.

pacman::p_load(R.utils, data.table, here, stringr, testit, dplyr)
dataset_name = "230731IMT"

## raw.info.csv is the .xlsx file that we get from genomecenter alongside fastq files
raw.info <- fread(here::here(paste0("data/", dataset_name, "/metadata/raw.info.csv")))
gex <- raw.info 
setDT(gex)
gex[, r := str_extract(file.name, pattern = "R[1-3]{1}(?=.fastq.gz)")]
gex <- gex[, .(name, file.name, r)]

## remove unnecessary extension to the sample name
gex[, `:=`(sample = paste(gsub(pattern = "^K_DNA_", replacement = "", gsub(pattern = "_[0-9]{2}", replacement = "", name)), r, sep = "_"), 
           fastq = here::here(paste0("data/", dataset_name, "/raw/", file.name)))]

gex.ret <- gex[file.exists(fastq)][, .(sample, fastq)]
fwrite(gex.ret, here::here(paste0("data/", dataset_name, "/metadata/sample.sheet.csv")))
fwrite(gex[, .(sample, file.name)], here::here(paste0("data/", dataset_name, "/metadata/renaming.csv")), sep = "\t", col.names = F)

here is a typical raw.info.csv file:

Disclaimer

The general structure of the RNA and DNA pipelines are similar, though the RNA pipeline hasn't been updated for a while, so there might be some idiosyncrasies between them, in terms of file name conventions.

Inside pipelines/genome.nf, there are process with which you can generate index files from the genome for kallisto and bismark, seperately.

Contact

Onurcan Bektas: onurcan.bektas@lmu.de