使用snakemake合并多个vcf文件

Question

我正在尝试使用snakemake通过染色体合并多个vcf文件。我的文件是这样的，如您所见，它具有各种坐标。合并所有chr1A和所有chr1B的最佳方法是什么？

chr1A:0-2096.filtered.vcf
chr1A:2096-7896.filtered.vcf
chr1B:0-3456.filtered.vcf
chr1B:3456-8796.filtered.vcf

我的伪代码：

chromosomes=["chr1A","chr1B"]

rule all:
    input: 
        expand("{sample}.vcf", sample=chromosomes)

rule merge:
    input:
        I1="path/to/file/{sample}.xxx.filtered.vcf",
        I2="path/to/file/{sample}.xxx.filtered.vcf",
    output:
        outf ="{sample}.vcf"
    shell:
        """
        java -jar picard.jar GatherVcfs I={input.I1} I={input.I2} O={output.outf}
        """

Answer 1

我没有snakemake atm，而且bioconda会永久使用XD，但将在星期一进行更新：

d = {"chr1A": ["0-2096.filtered.vcf", "2096-7896.filtered.vcf"]
"chr1B":["0-3456.filtered.vcf", "3456-8796.filtered.vcf"]}

chromosomes=["chr1A","chr1B"]

rule all:
    input: 
        expand("{sample}.vcf", sample=chromosomes)

rule merge:
    input:
        lambda w: d[w.chromosome]
    output:
        outf ="{chromosome}.vcf"
    params:
        lambda w: "I= " + "I=".join(d[w.chromosome])
    shell:
        """
        java -jar picard.jar GatherVcfs {params[0]} O={output.outf}
        """