Running the annotation step does not work with the full samplesheet file described on the website or with the variantcalling.csv file generated by the pipeline in a previous step

[MolPath-Bioinfo]

Description of the bug

Running only the annotation step using the csv file generated by the Sarek 3.1.2 pipeline in a previous run (variantcaller.csv) ends with the following error:

The sample-sheet only contains normal-samples, but the following tools, which were requested with "--tools", expect at least one tumor-sample : mutect2

The content of the variantcaller.csv corresponds to the full samplesheet described on the website and has the following format:

patient,sample,variantcaller,vcf
P1,H-2022-09665,mutect2,H-2022-09665.mutect2.filtered.vcf.gz
P1,H-2022-09665,strelka,H-2022-09665.strelka.variants.vcf.gz

Solution: adding the columns sex and status to the input file (directly after the patient column) solves the problem

patient,sex,status.sample,variantcaller,vcf
P1,XX,1,H-2022-09665,mutect2,H-2022-09665.mutect2.filtered.vcf.gz
P1,XX,1,H-2022-09665,strelka,H-2022-09665.strelka.variants.vcf.gz

The documentation on the website needs to be adapted and the 2 columns added to the full samplesheet example.

This behavior is not consistent in all steps. If I run the pipeline from the beginning with the tools
--tools mutect2,strelka,vep
then the error does not occur.

The columns sex and status seems to be required in all cases when one runs the individual steps using the --step parameter. I encountered this also while running only the variant calling step.

Command used and terminal output

nextflow run /home/thiele/local_apps/nf-core-sarek-3.1.2/workflow -profile singularity -c "/mnt/Molpath/users/mihaela/projects/variant_calling/WES/220728_A01542_0030_AHNCCMDRXY_3Aug_TWIST/VIP_tumorOnly/config/nf_wes.config" -params-file "/mnt/Molpath/users/mihaela/projects/variant_calling/WES/220728_A01542_0030_AHNCCMDRXY_3Aug_TWIST/VIP_tumorOnly/config/nf_params.json" --step annotate --vep_species "homo_sapiens" --vep_version '106.1' --vep_include_fasta --vep_loftee --vep_spliceregion


The sample-sheet only contains normal-samples, but the following tools, which were requested with "--tools", expect at least one tumor-sample : mutect2

Relevant files

No response

System information

• System: Linux 5.4.0-136-generic Ubuntu
• nf-core/sarek 3.1.2
• executor: local
• hardware: Desktop
• nextflow version: 22.10.4 build 5836
• container engine: Singularity

[FriederikeHanssen]

See also #948

[FriederikeHanssen]

Hi Mihaela! Is this working when just status?

[FriederikeHanssen]

hm I am thinking about this issue. and it should probably fail: mutect2 is not an annotation tool, but it starts from --step annotate, so it should fail but for a different reason

[MolPath-Bioinfo]

Hi Friederike, no it doesn't. It requests that one provides the sex and if tumor or normal sample. Theoretically it requests the full input file as one would run the preprocessing and variant call step too.

…

________________________________ Von: FriederikeHanssen ***@***.***> Gesendet: Montag, 27. März 2023 17:25:46 An: nf-core/sarek Cc: Mihaela Martis-Thiele; Author Betreff: Re: [nf-core/sarek] Running the annotation step does not work with the full samplesheet file described on the website or with the variantcalling.csv file generated by the pipeline in a previous step (Issue #949) Hi Mihaela! Is this working when just status? — Reply to this email directly, view it on GitHub<#949 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AWZR5QRIQWZ2Z66SUUKQOODW6GWPVANCNFSM6AAAAAAVEJFIZA>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[MolPath-Bioinfo]

Yes, mutect is not an annotation tool, but in this step I don't call it. I just make use of its output, which I want to annotate....

…

________________________________ Von: FriederikeHanssen ***@***.***> Gesendet: Montag, 27. März 2023 17:31:44 An: nf-core/sarek Cc: Mihaela Martis-Thiele; Author Betreff: Re: [nf-core/sarek] Running the annotation step does not work with the full samplesheet file described on the website or with the variantcalling.csv file generated by the pipeline in a previous step (Issue #949) hm I am thinking about this issue. and it should probably fail: mutect2 is not an annotation tool, but it starts from --step annotate, so it should fail but for a different reason — Reply to this email directly, view it on GitHub<#949 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AWZR5QXTWOLLWOXBFVOG5YDW6GXGBANCNFSM6AAAAAAVEJFIZA>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[FriederikeHanssen]

Tool validation will now be skipped if --step annotateis run. However, if starting from annotation, there shouldn't be any need to set a variant caller