[User Story] Unify multiple alignment steps into a single one

[rannick]

Description of feature

Need

As a USER I would like to reduce turn around times without compromising the quality of the analyses.

As a BIOINFORMATICIAN I would like avoid computational expensive steps to be unnecessarily duplicated.

Suggested approach

Unify STAR alignment steps for Arriba and STARfusion using settings that are compatible with both tools (and StringTie), and allow for the detection of fusions (including DUX4::IGH).

Suggested parameters:

STAR \
    --genomeDir GENOMEDIR \
    --readFilesIn R1.fastq.gz R2.fastq.gz \
    --runThreadN 12 \
    --outFileNamePrefix OUTFILEPREFIX. \
    --sjdbGTFfile ANNOTATION.gtf \
    --outSAMattrRGline 'ID:CASE  'SM:CASE  \
    --outReadsUnmapped None  \
    --outSAMstrandField intronMotif \
    --chimOutJunctionFormat 1 \
    --twopassMode None \
    --outFilterMultimapNmax 50 \
    --chimMultimapNmax 50 \
    --quantMode GeneCounts \
    --outSAMunmapped Within \
    --readFilesCommand zcat  \
    --alignSJstitchMismatchNmax 5 -1 5 5 \
    --outSAMtype BAM SortedByCoordinate \
    --chimSegmentMin 10 \
    --peOverlapNbasesMin 10 \
    --alignSplicedMateMapLminOverLmate 0.5 \
    --chimJunctionOverhangMin 10 \
    --chimScoreJunctionNonGTAG 0 \
    --chimScoreDropMax 30 \
    --chimScoreSeparation 1  \
    --chimSegmentReadGapMax 3 \
    --chimOutType Junctions WithinBAM

[rannick]

Will close this one too #493