Snakefile

#   hutspot - a DNAseq variant calling pipeline
#   Copyright (C) 2017-2019, Leiden University Medical Center
#
#   This program is free software: you can redistribute it and/or modify
#   it under the terms of the GNU Affero General Public License as published by
#   the Free Software Foundation, either version 3 of the License, or
#   (at your option) any later version.
#
#   This program is distributed in the hope that it will be useful,
#   but WITHOUT ANY WARRANTY; without even the implied warranty of
#   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#   GNU Affero General Public License for more details.
#
#   You should have received a copy of the GNU Affero General Public License
#   along with this program.  If not, see <https://www.gnu.org/licenses/>.
"""
Main Snakefile for the pipeline.

:copyright: (c) 2017-2019 Sander Bollen
:copyright: (c) 2017-2019 Leiden University Medical Center
:license: AGPL-3.0
"""

import json
import jsonschema
from functools import partial
from os.path import join, basename
from pathlib import Path
import itertools

# Read the json schema
with open(srcdir('config/schema.json'), 'rt') as fin:
    schema = json.load(fin)

# Validate the config against the schema
try:
    jsonschema.validate(config, schema)
except jsonschema.ValidationError as e:
    raise jsonschema.ValidationError(f'Invalid --configfile: {e.message}')

# If you specify a baitsfile, you also have to specify a targets file for
# picard
if "baitsfile" in config and "targetsfile" not in config:
    msg = 'Invalid --configfile: "baitsfile" specified without "targetsfile"'
    raise jsonschema.ValidationError(msg)

# A sample name cannot be a substring of another sample, since that breaks picard
# metrics parsing by multiqc
msg = 'Invalid --configfile: sample names should not overlap ("{s1}" is contained in "{s2}")'
for s1, s2 in itertools.permutations(config['samples'], 2):
    if s1 in s2:
        raise jsonschema.ValidationError(msg.format(s1=s1, s2=s2))

# Set default values
def set_default(key, value):
    """Set default config values"""
    if key not in config:
        config[key] = value

# Set the default config
set_default('scatter_size', 1000000000)
set_default('female_threshold', 0.6)

# Set the script paths
set_default("covstats", srcdir("src/covstats.py"))
set_default("collect_stats", srcdir("src/collect_stats.py"))
set_default("merge_stats", srcdir("src/merge_stats.py"))
set_default("stats_to_tsv", srcdir("src/stats_to_tsv.py"))
set_default("py_wordcount", srcdir("src/pywc.py"))
set_default("cutadapt_summary", srcdir("src/cutadapt_summary.py"))

containers = {
    "bcftools": "docker://quay.io/biocontainers/bcftools:1.9--ha228f0b_4",
    "bedtools-2.26-python-2.7": "docker://quay.io/biocontainers/mulled-v2-3251e6c49d800268f0bc575f28045ab4e69475a6:4ce073b219b6dabb79d154762a9b67728c357edb-0",
    "biopet-scatterregions": "docker://quay.io/biocontainers/biopet-scatterregions:0.2--0",
    "bwa-0.7.17-picard-2.18.7": "docker://quay.io/biocontainers/mulled-v2-002f51ea92721407ef440b921fb5940f424be842:43ec6124f9f4f875515f9548733b8b4e5fed9aa6-0",
    "cutadapt": "docker://quay.io/biocontainers/cutadapt:2.9--py37h516909a_0",
    "debian": "docker://debian:buster-slim",
    "fastqc": "docker://quay.io/biocontainers/fastqc:0.11.7--4",
    "gatk": "docker://broadinstitute/gatk3:3.7-0",
    "multiqc": "docker://quay.io/biocontainers/multiqc:1.8--py_2",
    "picard": "docker://quay.io/biocontainers/picard:2.22.8--0",
    "python3": "docker://python:3.6-slim",
    "samtools-1.7-python-3.6": "docker://quay.io/biocontainers/mulled-v2-eb9e7907c7a753917c1e4d7a64384c047429618a:1abf1824431ec057c7d41be6f0c40e24843acde4-0",
    "vtools": "docker://quay.io/biocontainers/vtools:1.0.0--py37h3010b51_0"
}

def get_readgroup(wildcards):
    return config["samples"][wildcards.sample]["read_groups"]

def get_readgroup_per_sample():
    for sample in config["samples"]:
        for rg in config["samples"][sample]["read_groups"]:
            yield rg, sample

def coverage_stats(wildcards):
    files = expand("{sample}/coverage/refFlat_coverage.tsv",
                   sample=config["samples"])
    return files if "refflat" in config else []

rule all:
    input:
        multiqc = "multiqc_report/multiqc_report.html",
        stats = "stats.json",
        stats_tsv = "stats.tsv",

        bam = expand("{sample}/bams/{sample}.bam",
                     sample=config["samples"]),

        vcfs = expand("{sample}/vcf/{sample}.vcf.gz",
                      sample=config["samples"]),

        vcf_tbi = expand("{sample}/vcf/{sample}.vcf.gz.tbi",
                         sample=config["samples"]),

        gvcfs = expand("{sample}/vcf/{sample}.g.vcf.gz",
                       sample=config["samples"]),

        gvcf_tbi = expand("{sample}/vcf/{sample}.g.vcf.gz.tbi",
                          sample=config["samples"]),

        fastqc_raw = (f"{sample}/pre_process/raw-{sample}-{read_group}/"
                      for read_group, sample in get_readgroup_per_sample()),

        fastqc_trim = (f"{sample}/pre_process/trimmed-{sample}-{read_group}/"
                      for read_group, sample in get_readgroup_per_sample()),

        cutadapt = (f"{sample}/pre_process/{sample}-{read_group}.txt"
                    for read_group, sample in get_readgroup_per_sample()),
        coverage_stats = coverage_stats,

rule create_markdup_tmp:
    """Create tmp directory for mark duplicates"""
    output: directory("tmp")
    container: containers["debian"]
    shell: "mkdir -p {output}"

rule genome:
    """Create genome file as used by bedtools"""
    input: config["reference"]
    output: "current.genome"
    container: containers["debian"]
    shell: "awk -v OFS='\t' {{'print $1,$2'}} {input}.fai > {output}"

rule cutadapt:
    """Clip fastq files"""
    input:
        r1 = lambda wc: (config['samples'][wc.sample]['read_groups'][wc.read_group]['R1']),
        r2 = lambda wc: (config['samples'][wc.sample]['read_groups'][wc.read_group]['R2']),
    output:
        r1 = "{sample}/pre_process/{sample}-{read_group}_R1.fastq.gz",
        r2 = "{sample}/pre_process/{sample}-{read_group}_R2.fastq.gz",
    log:
        "{sample}/pre_process/{sample}-{read_group}.txt"
    container: containers["cutadapt"]
    shell: "cutadapt -a AGATCGGAAGAG -A AGATCGGAAGAG "
           "--minimum-length 1 --quality-cutoff=20,20 "
           "--output {output.r1} --paired-output {output.r2} -Z "
           "{input.r1} {input.r2} "
           "--report=minimal > {log}"

rule align:
    """Align fastq files"""
    input:
        r1 = rules.cutadapt.output.r1,
        r2 = rules.cutadapt.output.r2,
        ref = config["reference"],
        tmp = ancient("tmp")
    params:
        rg = "@RG\\tID:{sample}-library-{read_group}\\tSM:{sample}\\tLB:library\\tPL:ILLUMINA"
    output: "{sample}/bams/{sample}-{read_group}.sorted.bam"
    container: containers["bwa-0.7.17-picard-2.18.7"]
    shell: "bwa mem -t 8 -R '{params.rg}' {input.ref} {input.r1} {input.r2} "
           "| picard -Xmx4G -Djava.io.tmpdir={input.tmp} SortSam "
           "CREATE_INDEX=TRUE TMP_DIR={input.tmp} "
           "INPUT=/dev/stdin OUTPUT={output} SORT_ORDER=coordinate"

def markdup_bam_input(wildcards):
    """Generate the INPUT for each bam file """
    return ["INPUT={sample}/bams/{sample}-{read_group}.sorted.bam".format(
                sample=wildcards.sample, read_group=rg)
            for rg in get_readgroup(wildcards)]

rule markdup:
    """Mark duplicates in BAM file"""
    input:
        bam = lambda wildcards:
        ("{sample}/bams/{sample}-{read_group}.sorted.bam".format(
            sample=wildcards.sample, read_group=rg)
            for rg in get_readgroup(wildcards)),
        tmp = ancient("tmp")
    output:
        bam = "{sample}/bams/{sample}.bam",
        bai = "{sample}/bams/{sample}.bai",
        metrics = "{sample}/bams/{sample}.metrics"
    params:
        bams=markdup_bam_input
    container: containers["picard"]
    shell: "picard -Xmx4G -Djava.io.tmpdir={input.tmp} MarkDuplicates "
           "CREATE_INDEX=TRUE TMP_DIR={input.tmp} "
           "{params.bams} OUTPUT={output.bam} "
           "METRICS_FILE={output.metrics} "
           "MAX_FILE_HANDLES_FOR_READ_ENDS_MAP=500"

rule bai:
    """Copy bai files as some genome browsers can only .bam.bai files"""
    input:
        bai = rules.markdup.output.bai
    output:
        bai = "{sample}/bams/{sample}.bam.bai"
    container: containers["debian"]
    shell: "cp {input.bai} {output.bai}"

def bqsr_bam_input(wildcards):
    """Generate the bam input string for each read group for BQSR"""
    template = "-I {sample}/bams/{sample}-{read_group}.sorted.bam"
    return " ".join([template.format(sample=wildcards.sample,
            read_group=rg) for rg in get_readgroup(wildcards)])

rule baserecal:
    """Base recalibrated BAM files"""
    input:
        bam = lambda wildcards:
        ("{sample}/bams/{sample}-{read_group}.sorted.bam".format(
            sample=wildcards.sample, read_group=rg)
            for rg in get_readgroup(wildcards)),
        ref = config["reference"],
        vcfs = config["known_sites"]
    output: "{sample}/bams/{sample}.baserecal.grp"
    params:
        known_sites = " ".join(
                expand("-knownSites {vcf}", vcf=config["known_sites"])
        ),
        bams = bqsr_bam_input
    container: containers["gatk"]
    shell: "java -XX:ParallelGCThreads=1 -jar /usr/GenomeAnalysisTK.jar -T "
           "BaseRecalibrator {params.bams} -o {output} -nct 8 "
           "-R {input.ref} -cov ReadGroupCovariate -cov QualityScoreCovariate "
           "-cov CycleCovariate -cov ContextCovariate {params.known_sites}"

checkpoint scatterregions:
    """Scatter the reference genome"""
    input:
        ref = config["reference"],
    params:
        size = config['scatter_size']
    output:
        directory("scatter")
    container: containers["biopet-scatterregions"]
    shell: "mkdir -p {output} && "
           "biopet-scatterregions -Xmx24G "
           "--referenceFasta {input.ref} --scatterSize {params.size} "
           "--outputDir scatter"

rule gvcf_scatter:
    """Run HaplotypeCaller in GVCF mode by chunk"""
    input:
        bam = rules.markdup.output.bam,
        bqsr = rules.baserecal.output,
        dbsnp = config["dbsnp"],
        ref = config["reference"],
        region = "scatter/scatter-{chunk}.bed"
    output:
        gvcf = temp("{sample}/vcf/{sample}.{chunk}.g.vcf.gz"),
        gvcf_tbi = temp("{sample}/vcf/{sample}.{chunk}.g.vcf.gz.tbi")
    wildcard_constraints:
        chunk = "[0-9]+"
    container: containers["gatk"]
    shell: "java -jar -Xmx4G -XX:ParallelGCThreads=1 /usr/GenomeAnalysisTK.jar "
           "-T HaplotypeCaller -ERC GVCF -I "
           "{input.bam} -R {input.ref} -D {input.dbsnp} "
           "-L '{input.region}' -o '{output.gvcf}' "
           "-variant_index_type LINEAR -variant_index_parameter 128000 "
           "-BQSR {input.bqsr} "
           "--GVCFGQBands 20 --GVCFGQBands 40 --GVCFGQBands 60 "
           "--GVCFGQBands 80 --GVCFGQBands 100 "

def aggregate_gvcf(wildcards):
    checkpoint_output = checkpoints.scatterregions.get(**wildcards).output[0]
    return expand("{{sample}}/vcf/{{sample}}.{i}.g.vcf.gz",
       i=glob_wildcards(os.path.join(checkpoint_output, 'scatter-{i}.bed')).i)

def aggregate_gvcf_tbi(wildcards):
    checkpoint_output = checkpoints.scatterregions.get(**wildcards).output[0]
    return expand("{{sample}}/vcf/{{sample}}.{i}.g.vcf.gz.tbi",
       i=glob_wildcards(os.path.join(checkpoint_output, 'scatter-{i}.bed')).i)

rule gvcf_gather:
    """Join the gvcf files together"""
    input:
        gvcfs = aggregate_gvcf,
        tbis = aggregate_gvcf_tbi,
    output:
        gvcf = "{sample}/vcf/{sample}.g.vcf.gz",
        gvcf_tbi = "{sample}/vcf/{sample}.g.vcf.gz.tbi"
    container: containers["bcftools"]
    shell: "bcftools concat {input.gvcfs} --allow-overlaps "
           "--output {output.gvcf} --output-type z && "
           "bcftools index --tbi --output-file {output.gvcf_tbi} {output.gvcf}"

rule genotype_scatter:
    """Run GATK's GenotypeGVCFs by chunk"""
    input:
        gvcf = rules.gvcf_scatter.output.gvcf,
        tbi = rules.gvcf_scatter.output.gvcf_tbi,
        ref= config["reference"]
    output:
        vcf = temp("{sample}/vcf/{sample}.{chunk}.vcf.gz"),
        vcf_tbi = temp("{sample}/vcf/{sample}.{chunk}.vcf.gz.tbi")
    wildcard_constraints:
        chunk = "[0-9]+"
    container: containers["gatk"]
    shell: "java -jar -Xmx15G -XX:ParallelGCThreads=1 "
           "/usr/GenomeAnalysisTK.jar -T "
           "GenotypeGVCFs -R {input.ref} "
           "-V {input.gvcf} -o '{output.vcf}'"

def aggregate_vcf(wildcards):
    checkpoint_output = checkpoints.scatterregions.get(**wildcards).output[0]
    return expand("{{sample}}/vcf/{{sample}}.{i}.vcf.gz",
       i=glob_wildcards(os.path.join(checkpoint_output, 'scatter-{i}.bed')).i)

def aggregate_vcf_tbi(wildcards):
    checkpoint_output = checkpoints.scatterregions.get(**wildcards).output[0]
    return expand("{{sample}}/vcf/{{sample}}.{i}.vcf.gz.tbi",
       i=glob_wildcards(os.path.join(checkpoint_output, 'scatter-{i}.bed')).i)

rule genotype_gather:
    """Gather all genotyping VCFs"""
    input:
        vcfs = aggregate_vcf,
        vcfs_tbi = aggregate_vcf_tbi
    output:
        vcf = "{sample}/vcf/{sample}.vcf.gz",
        vcf_tbi = "{sample}/vcf/{sample}.vcf.gz.tbi"
    container: containers["bcftools"]
    shell: "bcftools concat {input.vcfs} --allow-overlaps "
           "--output {output.vcf} --output-type z && "
           "bcftools index --tbi --output-file {output.vcf_tbi} {output.vcf}"

## bam metrics
rule mapped_reads_bases:
    """Calculate number of mapped reads"""
    input:
        bam = rules.markdup.output.bam,
        pywc = config["py_wordcount"]
    output:
        reads = "{sample}/bams/{sample}.mapped.num",
        bases = "{sample}/bams/{sample}.mapped.basenum"
    container: containers["samtools-1.7-python-3.6"]
    shell: "samtools view -F 4 {input.bam} | cut -f 10 | python {input.pywc} "
           "--reads {output.reads} --bases {output.bases}"

rule unique_reads_bases:
    """Calculate number of unique reads"""
    input:
        bam = rules.markdup.output.bam,
        pywc = config["py_wordcount"]
    output:
        reads = "{sample}/bams/{sample}.unique.num",
        bases = "{sample}/bams/{sample}.usable.basenum"
    container: containers["samtools-1.7-python-3.6"]
    shell: "samtools view -F 4 -F 1024 {input.bam} | cut -f 10 | "
           "python {input.pywc} --reads {output.reads} --bases {output.bases}"

## fastqc
rule fastqc_raw:
    """Run fastqc on raw fastq files"""
    input:
        r1 = lambda wc: (config['samples'][wc.sample]['read_groups'][wc.read_group]['R1']),
        r2 = lambda wc: (config['samples'][wc.sample]['read_groups'][wc.read_group]['R2']),
    output:
        directory("{sample}/pre_process/raw-{sample}-{read_group}/")
    container: containers["fastqc"]
    shell: "fastqc --threads 4 --nogroup -o {output} {input.r1} {input.r2} "

rule fastqc_postqc:
    """Run fastqc on fastq files post pre-processing"""
    input:
        r1 = rules.cutadapt.output.r1,
        r2 = rules.cutadapt.output.r2
    output:
        directory("{sample}/pre_process/trimmed-{sample}-{read_group}/")
    container: containers["fastqc"]
    shell: "fastqc --threads 4 --nogroup -o {output} {input.r1} {input.r2} "

## coverage
rule covstats:
    """Calculate coverage statistics on bam file"""
    input:
        bam = rules.markdup.output.bam,
        genome = "current.genome",
        covpy = config["covstats"],
        bed = config.get("targetsfile","")
    params:
        subt = "Sample {sample}"
    output:
        covj = "{sample}/coverage/covstats.json",
        covp = "{sample}/coverage/covstats.png"
    container: containers["bedtools-2.26-python-2.7"]
    shell: "bedtools coverage -sorted -g {input.genome} -a {input.bed} "
           "-b {input.bam} -d  | python {input.covpy} - --plot {output.covp} "
           "--title 'Targets coverage' --subtitle '{params.subt}' "
           "> {output.covj}"

rule vtools_coverage:
    """Calculate coverage statistics per transcript"""
    input:
        gvcf = rules.gvcf_gather.output.gvcf,
        tbi = rules.gvcf_gather.output.gvcf_tbi,
        ref = config.get('refflat', "")
    output:
        tsv = "{sample}/coverage/refFlat_coverage.tsv"
    container: containers["vtools"]
    shell: "vtools-gcoverage -I {input.gvcf} -R {input.ref} > {output.tsv}"

rule collect_cutadapt_summary:
    """Colect cutadapt summary from each readgroup per sample """
    input:
        cutadapt = lambda wildcards:
        ("{sample}/pre_process/{sample}-{read_group}.txt".format(
            sample=wildcards.sample, read_group=read_group)
            for read_group in get_readgroup(wildcards)),
        cutadapt_summary= config["cutadapt_summary"]
    output: "{sample}/cutadapt.json"
    container: containers["python3"]
    shell: "python {input.cutadapt_summary} --sample {wildcards.sample} "
           "--cutadapt-summary {input.cutadapt} > {output}"

rule collectstats:
    """Collect all stats for a particular sample"""
    input:
        mnum = rules.mapped_reads_bases.output.reads,
        mbnum = rules.mapped_reads_bases.output.bases,
        unum = rules.unique_reads_bases.output.reads,
        ubnum = rules.unique_reads_bases.output.bases,
        cov = rules.covstats.output.covj if "targetsfile" in config else [],
        cutadapt = rules.collect_cutadapt_summary.output,
        colpy = config["collect_stats"]
    params:
        fthresh = config["female_threshold"]
    output: "{sample}/{sample}.stats.json"
    container: containers["python3"]
    shell: "python {input.colpy} --sample-name {wildcards.sample} "
           "--mapped-num {input.mnum} --mapped-basenum {input.mbnum} "
           "--unique-num {input.unum} --usable-basenum {input.ubnum} "
           "--female-threshold {params.fthresh} "
           "--cutadapt {input.cutadapt} "
           "--covstats {input.cov} > {output}"

rule multiple_metrics:
    """Run picard CollectMultipleMetrics"""
    input:
        bam = rules.markdup.output.bam,
        ref = config["reference"],
    params:
        prefix = "{sample}/bams/{sample}",
    output:
        alignment = "{sample}/bams/{sample}.alignment_summary_metrics",
        insertMetrics = "{sample}/bams/{sample}.insert_size_metrics"
    container: containers["picard"]
    shell: "picard CollectMultipleMetrics "
           "I={input.bam} O={params.prefix} "
           "R={input.ref} "
           "PROGRAM=CollectAlignmentSummaryMetrics "
           "PROGRAM=CollectInsertSizeMetrics "

rule bed_to_interval:
    """Run picard BedToIntervalList

    This is needed to convert the bed files for the capture kit to a format
    picard can read
    """
    input:
        targets = config.get("targetsfile",""),
        baits = config.get("baitsfile",""),
        ref = config["reference"]
    output:
        target_interval = "target.interval",
        baits_interval = "baits.interval"
    output:
    container: containers["picard"]
    shell: "picard BedToIntervalList "
            "I={input.targets} O={output.target_interval} "
            "SD={input.ref} && "
            "picard BedToIntervalList "
            "I={input.baits} O={output.baits_interval} "
            "SD={input.ref} "

rule hs_metrics:
    """Run picard CollectHsMetrics"""
    input:
        bam = rules.markdup.output.bam,
        ref = config["reference"],
        targets = rules.bed_to_interval.output.target_interval,
        baits = rules.bed_to_interval.output.baits_interval,
    output: "{sample}/bams/{sample}.hs_metrics.txt"
    container: containers["picard"]
    shell: "picard CollectHsMetrics "
           "I={input.bam} O={output} "
           "R={input.ref} "
           "BAIT_INTERVALS={input.baits} "
           "TARGET_INTERVALS={input.targets}"

rule multiqc:
    """
    Create multiQC report
    Depends on stats.tsv to forcefully run at end of pipeline
    """
    input:
        bam = expand("{sample}/bams/{sample}.bam", sample=config["samples"]),
        metric = expand("{sample}/bams/{sample}.metrics",
                        sample=config["samples"]),
        alignment_metrics = expand(
                "{sample}/bams/{sample}.alignment_summary_metrics",
                sample=config["samples"]
        ),
        insert_metrics = expand(
                "{sample}/bams/{sample}.insert_size_metrics",
                sample=config["samples"]
        ),
        fastqc_raw = (f"{sample}/pre_process/raw-{sample}-{read_group}/"
                      for read_group, sample in get_readgroup_per_sample()),

        fastqc_trim = (f"{sample}/pre_process/trimmed-{sample}-{read_group}/"
                      for read_group, sample in get_readgroup_per_sample()),

        hs_metric = expand("{sample}/bams/{sample}.hs_metrics.txt",
                           sample=config["samples"]) if "baitsfile" in config else []

    output:
        html = "multiqc_report/multiqc_report.html",
        insertSize = "multiqc_report/multiqc_data/multiqc_picard_insertSize.json",
        AlignmentMetrics = "multiqc_report/multiqc_data/multiqc_picard_AlignmentSummaryMetrics.json",
        DuplicationMetrics = "multiqc_report/multiqc_data/multiqc_picard_dups.json",
        HsMetrics = "multiqc_report/multiqc_data/multiqc_picard_HsMetrics.json" if "baitsfile" in config else []
    container: containers["multiqc"]
    shell: "multiqc --data-format json --force --outdir multiqc_report . "
           "|| touch {output}"

rule merge_stats:
    """Merge all stats of all samples"""
    input:
        cols = expand("{sample}/{sample}.stats.json",
                      sample=config['samples']),
        mpy = config["merge_stats"],
        insertSize = rules.multiqc.output.insertSize,
        AlignmentMetrics = rules.multiqc.output.AlignmentMetrics,
        DuplicationMetrics = rules.multiqc.output.DuplicationMetrics,
        HsMetrics = rules.multiqc.output.HsMetrics
    output: "stats.json"
    container: containers["vtools"]
    shell: "python {input.mpy} --collectstats {input.cols} "
           "--picard-insertSize {input.insertSize} "
           "--picard-AlignmentMetrics {input.AlignmentMetrics} "
           "--picard-DuplicationMetrics {input.DuplicationMetrics} "
           "--picard-HsMetrics {input.HsMetrics} > {output}"

rule stats_tsv:
    """Convert stats.json to tsv"""
    input:
        stats = rules.merge_stats.output,
        sc = config["stats_to_tsv"]
    output: "stats.tsv"
    container: containers["python3"]
    shell: "python {input.sc} -i {input.stats} > {output}"