Pomfrey

Hematology Twist Pipeline

To run the pipeline you need Snakemake and Singularitys installed. At Uppsala it is used together with slurm-drmaa to submit on the local HPC. If Horizon Myeloid DNA Reference Standard is used it should be named HD829 to be processed separately and not hold up the pipeline.

Files & Caches

• SampleSheetUsed.csv: A csv-file with produced when demultiplexing on Illumina machine. Is used to order samples in MultiQC table. The script only use the column Sample_Name. Lines needed are in the file except the actual sample-lines are:

  [Data]
  Sample_ID,Sample_Name,Description,index,I7_Index_ID,index2,I5_Index_ID,Sample_Project

• Bedfile: Four columns: chr, start, stop, regionname. No header. The bed-file is used as both target and bait intervals in picard HsMetric for bases on target stats. The fourth column is used in CARTools to identify low coverage regions.
• Pindel bedfile: Since pindel is quite slow a smaller bedfile with limited regions is needed to run pindel.
• Intervals-file: Corresponding to the bedfile for picard jobs. Can be generated in GATK4 with: gatk fileBedToIntervalList --INPUT $bedfile --OUTPUT $interval_output --SEQUENCE_DICTIONARY $reference_dict
• Reference: Fasta refernce index both with bwa index and .fai.
• Artefact file: Known artefacts in bedformat
• Germline file:
• Pindel artefact file: Known pindel artefacts, chr, pos, gene.
• COSMIC hemato counts: COSMIC file with number of hemato hits. Columns (tab seperated): GENE_NAME, ACCESSION_NUMBER, GENE_CDS_LENGTH, HGNC_ID PRIMARY_SITE, PRIMARY_HISTOLOGY, GENOMIC_MUTATION_ID, LEGACY_MUTATION_ID, MUTATION_ID, MUTATION_CDS, MUTATION_AA, MUTATION_GENOME_POSITION, MUTATION_STRAND, SNP, MUTATION_SOMATIC_STATUS, N_observations, Chromosome, chr, Start, End, Build
• Hotspot list: List of regions where higher coverage is important. Each region and its coverage is listed in the sheet "HotSpot" in the xlsx-file
• VEP cache: Need to download cache for vep to run. Read more on the different versions of vep-cache. singularity exec --bind $PWD vep-container.simg perl /opt/vep/src/ensembl-vep/INSTALL.pl -s homo_sapiens_refseq --CACHEDIR vep-data-99.0/ -a c --ASSEMBLY GRCh37

Singularity Containers

Program	Version	Source
Bcftools	1.9	docker://quay.io/biocontainers/bcftools:1.9--h68d8f2e_8
Cutadapt	2.5	docker://quay.io/biocontainers/cutadapt:2.5--py37h516909a_0
Fastqc	0.11.8	docker://quay.io/biocontainers/fastqc:0.11.8--1
Freebayes	1.3.1	docker://quay.io/biocontainers/freebayes:1.3.1--py37h56106d0_0
GATK4	4.1.7.0	docker://broadinstitute/gatk:4.1.7.0
MultiQC	1.7	docker://quay.io/biocontainers/multiqc:1.7--py_3
Pindel	0.2.58	docker://shuangbroad/pindel:v0.2.5b8
Vardict-java	1.7.0	docker://quay.io/biocontainers/vardict-java:1.7.0--0
Vep	99	docker://ensemblorg/ensembl-vep:release_99.0
Vt	0.57721	docker://quay.io/biocontainers/vt:0.57721--hdf88d34_2

Program	Comment	Source
bcbio-ensemble-recall		bcbio-variation-recall.simg
Bwa & Samtools & Picard	v.7.17 & v.1.9 & v.2.20.1	bwa-snakemakewrapper.def
CARTool	In-house program avaialble at Github	CARTools-200206.simg
Igv v.2.4.10 & xvfb	Not working to load hg19 genome into container	igv.def
Pisces v.5.2.11.163	Microsoft dotnet v.2.1	Pisces_singularity.def
Python3	Libraries: csv, pysam, xlsxwriter,date, itemgetter, yaml	python-pysam.def

Config Files

Samples Config

Yaml config file with samples to process and local variables and files for the pipeline. Configfile name has to be ${sequencerun}_config.yaml

programdir:
  dir: "${PATH_TO_POMFREY}"

reference:
    ref: "" #Path to fasta ref and .fai
    bwa: "" #Path to bwa indexed ref

configCache:
    multiqc: "${PATH_TO_POMFREY}/src/report/multiqc_config.yaml"
    vep: "" #Path to downloaded VEP cache
    hemato: "" #Path to COSMIC file for number of hematology hits. Downloaded from COSMIC site..
    variantlist: "" #Path to file where all variants are written to later create artefact and germlinefilter files.

bed:
    bedfile: "" #Path to main bedfile
    intervals: "" Path to interval file corresponding to main bedfile
    pindel: ""  #Path to pindel bedfile
    indelartefact: "" # Path to indel artefacts since all indels from vardict and mutect2 are included
    pindelArtefact: ""  #Path to pindel artefact file
    cartool: "" #Path to (main) bedfile or different if interested in the coverage of different regions.
    hotspot: "" #Path to hotpspotlist
    artefact: "" #Path to artefact filter file
    germline: "" #Path to germline filter file

CNV:
    PoN: "" # Path to you readCountPoN.hdf5 file for the panel of normals
    bedPoN: "" # Bedfile used for PoN, could be normal bedfile
    interval: "" #interval list
    cyto: "" # Translation from cyto coordinate to genomic position [chr start end cytoCoord]

singularitys:
    cutadapt: ""
    bwa: ""
    fastqc: ""
    cartool: ""
    bcftools: ""
    freebayes: ""
    pisces: ""
    vardict: ""
    gatk4: ""
    pindel: ""
    vep: ""
    recall: ""
    python: ""
    vt: ""
    igv: ""
    multiqc: ""

cartool:
    cov: "100 200 1000" #Coverage limits, first number minCov, second for hotspotlist, third wishful

methods:   # The (trust) order of vcfs into ensemble recall
    mutect2: "mutect2"
    vardict: "vardict"
    pisces: "pisces"
    freebayes: "freebayes"

seqID:
    sequencerun: ""  

samples:
   "${sampleName}": "${path_to_R1_fastq_gz}"

Cluster Config

Json file with config for submission on HPC. Need to be specified to suit you HPC. See cluster-config.json for example.

Snakemake command

snakemake -p -j ${max_nr_jobs_submitted} --drmaa "-A ${project} -p core -t {cluster.time} -n {cluster.n} --nodes=1-1 " --use-singularity --cluster-config ${cluster_config} -s ${PATH_TO_POMFREY}/src/somaticPipeline.smk --singularity-args " --cleanenv --bind /data/ --bind /projects/ " --configfile ${sample_config}