Question

我正在编写一个Snakemake管道，并且有一些规则通过qsub运行。这些规则之一称为run_MSA，其代码为

rule run_MSA:
    input:
        file=model_dir + "/{projectID}_MODEL_{remove_lower_t}_{remove_higher_t}/{projectID}_start.fasta"
    output:
        alignment= model_dir + "/{projectID}_MODEL_{remove_lower_t}_{remove_higher_t}/align1.phy"
    threads: cluster_config['run_MSA']['threads']
    log:
        "logs/quantizyme_model_MSA/{projectID}_filtering_{remove_lower_t}_{remove_higher_t}.log"
    shell:
        """
        clustalo \
        --threads {threads} \
        -i {input} \
        -o {output.alignment} \
        --outfmt=phy  > {log} 2>&1
        """

cluster.yaml文件是

__default__:
    time: 1:00:00
    threads: 20

run_MSA:
    time: 1:00:00
    threads: 18

run_MSA_subtree:
    time: 1:00:00
    threads: 10

subtreeing2:
    time: 1:00:00
    threads: 15

我遇到了一些奇怪的事情：当我使用-n选项运行snakemake管道时，按照以下说明，一切似乎都很好：

snakemake -nrp \
--config projectID=ref_AA2 remove_lower_t=100 remove_higher_t=1500 remove_seqs=TRUE subtrees=3 \
-j 100 --cluster-config cluster.yaml \
--cluster 'qsub -V -l h_rt={cluster.time} -pe smp {cluster.threads} -cwd -j y' \
-s quantizyme_model.2.snakefile

    Building DAG of jobs...
    Job counts:
        count   jobs
        1   all
        1   clustering
        1   compress_out_folder
        1   run_MSA
        3   run_MSA_subtree
        3   subtreeing1
        3   subtreeing2
        13

    [Wed Oct 24 18:22:59 2018]
    rule run_MSA:
        input: models/ref_AA2_MODEL_100_1500/ref_AA2_start.fasta
        output: models/ref_AA2_MODEL_100_1500/align1.phy
        log: logs/quantizyme_model_MSA/ref_AA2_filtering_100_1500.log
        jobid: 12
        reason: Missing output files: models/ref_AA2_MODEL_100_1500/align1.phy
        wildcards: projectID=ref_AA2, remove_lower_t=100, remove_higher_t=1500
        threads: 18


            clustalo         --threads 18         -i models/ref_AA2_MODEL_100_1500/ref_AA2_start.fasta         -o models/ref_AA2_MODEL_100_1500/align1.phy         --outfmt=phy  > logs/quantizyme_model_MSA/ref_AA2_filtering_100_1500.log 2>&1

为简单起见，我仅报告规则run_MSA的标准输出。但是，当我在不使用-n选项的情况下运行同一命令行时，会出现此错误：

KeyError in line 65 of /nfs4/my-gridfront/mykopat-proj3/mykopat-metatrans/Quanty_test2/quantizyme_snakemake_test_3/quantizyme_model.2.snakefile:
'run_MSA'
File "/nfs4/my-gridfront/mykopat-proj3/mykopat-metatrans/Quanty_test2/quantizyme_snakemake_test_3/quantizyme_model.2.snakefile", line 65, in <module>

所以对我来说，当蛇形运行作为空运行运行时，似乎可以很好地读取相同的关键字，但是在实际执行过程中却不是。我想念什么？

非常感谢，

Domenico

在snakemake中使用qsub：KeyError

0 个答案: