如何在Snakemake中读取字典输入并解决NameError问题

问题场景 #

我将FastQ文件路径存储在字典中，尝试在Snakemake规则里访问这些路径，但运行代码时触发NameError。

我的代码：

infile_loc = {
    "Sample1": ["/home/joshua/snakemake_works/input/Sample1_R1.fq.gz", "/home/joshua/snakemake_works/input/Sample1_R2.fq.gz"],
    "Sample2": ["/home/joshua/snakemake_works/input/Sample2_R1.fq.gz", "/home/joshua/snakemake_works/input/Sample2_R2.fq.gz"]
}

SAMPLES = infile_loc.keys()

rule print_file_name:
    input:
        expand("{sample}", sample=SAMPLES)
    params:
        read1, read2 = lambda wildcards: infile_loc[wildcards.sample]
    script:
        """
        echo {params.read1}
        echo {params.read2}
        """

运行后报错：

NameError in file /home/joshua/snakemake_works/Snakefile, line 9:
name 'read1' is not defined
  File "/home/joshua/snakemake_works/Snakefile", line 9, in <module>

错误原因 #

Snakemake的params块不支持直接使用read1, read2 = lambda...这种变量解构语法。这里的read1和read2在当前上下文属于未定义变量，因此触发NameError。

两种解决方法 #

方法1：为每个参数单独编写lambda表达式

直接给read1和read2分别指定lambda，从字典的列表值中提取对应位置的路径：

infile_loc = {
    "Sample1": ["/home/joshua/snakemake_works/input/Sample1_R1.fq.gz", "/home/joshua/snakemake_works/input/Sample1_R2.fq.gz"],
    "Sample2": ["/home/joshua/snakemake_works/input/Sample2_R1.fq.gz", "/home/joshua/snakemake_works/input/Sample2_R2.fq.gz"]
}

SAMPLES = infile_loc.keys()

rule print_file_name:
    input:
        expand("{sample}", sample=SAMPLES)
    params:
        read1=lambda wildcards: infile_loc[wildcards.sample][0],
        read2=lambda wildcards: infile_loc[wildcards.sample][1]
    script:
        """
        echo {params.read1}
        echo {params.read2}
        """

方法2：让lambda返回字典映射参数

让lambda函数返回包含read1和read2键的字典，Snakemake会自动将字典键映射为params的属性：

infile_loc = {
    "Sample1": ["/home/joshua/snakemake_works/input/Sample1_R1.fq.gz", "/home/joshua/snakemake_works/input/Sample1_R2.fq.gz"],
    "Sample2": ["/home/joshua/snakemake_works/input/Sample2_R1.fq.gz", "/home/joshua/snakemake_works/input/Sample2_R2.fq.gz"]
}

SAMPLES = infile_loc.keys()

rule print_file_name:
    input:
        expand("{sample}", sample=SAMPLES)
    params:
        lambda wildcards: {
            "read1": infile_loc[wildcards.sample][0],
            "read2": infile_loc[wildcards.sample][1]
        }
    script:
        """
        echo {params.read1}
        echo {params.read2}
        """

额外优化提示 #

原代码中input: expand("{sample}", sample=SAMPLES)没有指向实际文件，若规则不需要依赖输入文件可直接删除；如果需要关联FastQ文件作为输入，可改为：

input:
    expand("/home/joshua/snakemake_works/input/{sample}_R{1,2}.fq.gz", sample=SAMPLES)

内容的提问来源于stack exchange，提问作者Joshua Jebaraj