DSL2 modules

[BarryDigby]

Description of feature

Categorization of the workflow at the process level with the corresponding modules needed to port to 'DSL2'. Once the modules have been created, I can place more shape on this in terms of subworkflows.

N.B: please checkout new branches for individual features and push to the DSL2 branch, not dev.

Input files

Currently, circRNA takes as input a samplesheet.csv file and a phenotype.csv file. Functions already exist to check these files, all that is needed is to place these in an input_check.nf local subworkflow.

I would like to incorporate strandedness like other nf-core workflows. Will check which circRNA quantification tools have a parameter denoting strandedness.

Pre-processing

The workflow takes as input fastq or bam files (which are converted to fastq using picard SamToFastq) and performs FastQC on the raw reads prior to trimming using BBDUK. The trimmed reads are then checked using FastQC again and placed in channels for downstream analyses.

• FastQC
• MultiQC
• BBDUK
• picard/SamToFastq (I don't care if we drop this functionality.)

circRNA Discovery

Several tools utilize the same aligner, there will be duplicates here.

CIRIquant

• bwa index
• hisat build
• ciriquant

CIRCexplorer2

• STAR genomegenerate
• STAR align (2 Pass mode)
• circexplorer2 parse
• circexplorer2 annotate

circRNA_finder

• star genomegenerate
• star align (2 Pass mode)
• circRNA_Finder (postProcessStarAlignment.pl script)

DCC

DCC maps paired-end reads jointly and separately using STAR 2 pass mode. The goal is to generate chimeric.junction.out files from joint STAR mapping and individual read 1 and read 2 STAR mapping.

• star genomegenerate
• star align (2 Pass)
• dcc

find_circ

• bowtie2 build
• bowtie2 align
• find_circ find_anchors
• find_circ find_circ

Mapsplice

• bowtie build
• mapsplice align
• circexplorer2 parse
• circexplorer2 annotate

Segemehl

• segemehl align

Custom scripts to parse segemehl output, no need to create a module.

circRNA annotation

customized bash script to standardise the annotation outputs from the seven quantification tools.

circRNA FASTA sequence

customized bash script to generate the mature spliced sequence in FASTA format, and append the back-splice junction sequence for miRNA target prediction.

circRNA count matrix

consolidate the circRNAs called by multiple tools on a per sample basis, generate the count matrix.

miRNA target prediction

miRanda

• miranda

TargetScan

• targetscan. biocontainers #475

custom script to amalgamate the results from both tools.

Differential expression

• hisat build
• hisat align
• stringite

Custom R scripts for DESeq2 and CircTest, no need to create modules.

[nictru]

I think all these modules have been properly implemented. Can we close this issue?