rnaseq_without_DGE-Mouse10.yml

---
global:
    # ROOT Directory configurations
    - indir:     data/processed
    - outdir:    data/analysis
    # Find Samples
    - sample_rule: (Sample.*)$
    - by_sample_outdir: 1
    - find_by_dir: 1
    # Analsis Dir Setup
    - analysis_dir: data/analysis/
    - raw_dir: data/raw
    - rename_dir: data/raw/rename
    - trimmomatic_dir: "data/processed/{$sample}/trimmomatic"
    - trimmomatic_fastqc_dir: "data/processed/{$sample}/trimmomatic_fastqc"
    - htseq_dir: "data/analysis/Counts"
    #The data_dirs and very site specific variables  - be sure to change these!
    - data_dir: "/scratch/Reference_Genomes/Public/Vertebrate_mammalian/Mus_musculus/ENSEMBL-release-82-GRCm38"
    - ANNOTATION:  "{$self->data_dir}/tr_Mus_musculus_GRCm38/Mus_musculus.GRCm38.82"
    - REFERENCE: "{$self->data_dir}/Mus_musculus.GRCm38.dna.toplevel"
    #HPC Directives
    - HPC:
       - module: 'gencore gencore_dev gencore_rnaseq'
       - partition: 'serial'
       - commands_per_node: 1
       - cpus_per_task: 1
rules:
    - create_analysis_dirs:
        local:
                - create_outdir: 0
                - override_process: 1
                - HPC:
                   - walltime: '02:00:00'
                   - mem: '4GB'
                   - commands_per_node: 100
        process: |
            mkdir -p {$self->analysis_dir}/FPKMs &&  \
            mkdir -p {$self->analysis_dir}/BAMs  && \
            mkdir -p {$self->analysis_dir}/Transcripts && \
            mkdir -p {$self->analysis_dir}/CUFFDIFF && \
            mkdir -p {$self->analysis_dir}/Counts/DESeq2
    - tophat2:
        local:
                - create_outdir: 1
                - indir: "{$self->trimmomatic_dir}"
                - HPC:
                   - deps: create_analysis_dirs
                   - walltime: '12:00:00'
                   - mem: '60GB'
                   - cpus_per_task: 12
        process: |
            #TASK tags={$sample}
            tophat2 -o {$self->outdir} \
            --no-novel-juncs -p 12 --transcriptome-index={$self->ANNOTATION} \
                {$self->{REFERENCE}} \
                {$self->trimmomatic_dir}/{$sample}_read1_trimmomatic_1PE.fastq.gz \
                {$self->trimmomatic_dir}/{$sample}_read2_trimmomatic_2PE.fastq.gz
    - cufflinks:
       ## Cufflinks will only output a file name accepted_hits.bam
       ## to a particular directory
       ## We leave it here for now and move it in the move_data dir
        local:
                - INPUT: "{$self->indir}/accepted_hits.bam"
                - OUTPUT: "{$self->analysis_dir}/BAMs/{$sample}_accepted_hits.bam"
                - HPC:
                  - deps: tophat2
                  - mem: '60GB'
                  - walltime: '10:00:00'
                  - cpus_per_task: 12
        process: |
                #TASK tags={$sample}
                cufflinks -p 12 --no-update-check -o {$self->outdir} \
                    -G {$self->ANNOTATION} {$self->INPUT}
    - move_data:
        ## We move data around here for downstream analysis
        ## Some tool downstream requires its data to be in the same directory
        local:
                - tophat2_dir: 'data/analysis/{$sample}/tophat2'
                - outdir: "{$self->analysis_dir}"
                - INPUT: "{$self->tophat2_dir}/accepted_hits.bam"
                - OUTPUT: "{$self->analysis_dir}/BAMs/{$sample}_accepted_hits.bam"
                - HPC:
                  - deps: cufflinks
                  - mem: '4GB'
                  - walltime: '10:00:00'
                  - cpus_per_task: 1
                  - commands_per_node: 100
        process: |
                #TASK tags={$sample}
                mv {$self->INPUT} {$self->OUTPUT}; \
                    mv {$self->outdir}/gene* {$self->analysis_dir}/FPKMs/{$sample}_gene_fpkms.txt;  \
                    mv {$self->outdir}/transcrip* {$self->analysis_dir}/Transcripts/{$sample}_transcripts.gtf
    - samtools_sort:
        local:
                - outdir: "{$self->analysis_dir}"
                - INPUT: "{$self->analysis_dir}/BAMs/{$sample}_accepted_hits.bam"
                - OUTPUT: "{$self->analysis_dir}/BAMs/{$sample}_sorted.accepted_hits.bam"
                - HPC:
                  - deps: move_data
                  - mem: '80GB'
                  - walltime: '10:00:00'
                  - cpus_per_task: 6
        process: |
                #TASK tags={$sample}
                samtools sort -n -@ 6 -O bam \
                    -o {$self->OUTPUT} \
                    {$self->INPUT}
    - htseq_count:
        ## HtSeq count, or at least this version, requires that files be named a certain way
        ## And in the same directory
        local:
                - INPUT: "{$self->analysis_dir}/BAMs/{$sample}_sorted.accepted_hits.bam"
                - OUTPUT: "{$self->analysis_dir}/Counts/htseq_$name.txt"
                - HPC:
                  - deps: samtools_sort
                  - mem: '35GB'
                  - walltime: '10:00:00'
        process: |
                #TASK tags={$sample}
                name=`echo {$sample}| sed s/Sample_//` && \
                htseq-count -f bam -s no -t exon \
                    -i gene_id \
                    {$self->INPUT} \
                    {$self->ANNOTATION} > {$self->OUTPUT} && \
                    sed -i '/^__.*/d'  {$self->OUTPUT}
    - clean_dirs:
        ## Since we moved data around we will clean up the directories
        local:
                - data_dir: 'data/ercc_analysis'
                - override_process: 1
                - HPC:
                  - deps: htseq_count
                  - mem: '4GB'
                  - walltime: '10:00:00'
        process: |
                find {$self->data_dir} -type d -empty -delete