Hi there,

First post, apologies for any social faux pas.

I'm working towards a BSc(Hons) in genetics, and working with RNA-seq data for CRC and liver metastastic tumours. No normal tissue samples, I'm afraid.

I'm currently just trying to merge VCF files (one from BCFTools, the other from Freebayes) of the same sample using gatk3's "CombineVariants", when I get the above error. I'm using Snakemake to help automate this process, and my current command for the merging rule is as follows:

rule merge_vcf:

Merge VCF files and uniquify novel calls via GATK3

input:
    bvcf = join(VARDIR, "{SAMPLE}.bcf.vcf"),
    fvcf = join(VARDIR, "{SAMPLE}.free.vcf")
output:
    mvcf = temp(join(VARDIR, "{SAMPLE}.merged.vcf"))
log:
    log1 = join(LOGDIR, "merged/{SAMPLE}.stdout"),
    log2 = join(LOGDIR, "merged/{SAMPLE}.stderr")
benchmark:
    join(BENCHDIR, "merged/{SAMPLE}.txt")
conda: "envs/gatk.yaml"
shell:
    """
    nice gatk \
    CombineVariants \
    -R {GENOME} \
    --variant {input.bvcf} \
    --variant {input.fvcf} \
    -o {output.mvcf} \
    -genotypeMergeOptions UNIQUIFY \
    > {log.log1} \
    2> {log.log2}
    """

I'm not sure what I've done wrong, as I based the command line from the GATK website (https://software.broadinstitute.org/gatk/documentation/tooldocs/3.8-0/org_broadinstitute_gatk_tools_walkers_variantutils_CombineVariants.php), with some tweaks to get it working on Snakemake.

Any help would be appreciated, and hopefully aid with the addition of other GATK tools in the future (such as those suggested in the RNA-seq GATK best practises).

Thanks in advance!