Error: --score variant ID '2:168745247:A:AT' appears multiple times in main dataset.

[asmaa-a-abdelwahab]

I got this error in the PLINK2_SCORE process. Here is the command log:

PLINK v2.00a3.3LM 64-bit Intel (3 Jun 2022)    www.cog-genomics.org/plink/2.0/
(C) 2005-2022 Shaun Purcell, Christopher Chang   GNU General Public License v3
Logging to Population_Report_ALL_additive_0.log.
Options in effect:
  --memory 204800
  --out Population_Report_ALL_additive_0
  --pfile vzs vcf_Population_Report_ALL
  --score Population-Report_ALL_additive_0.scorefile.gz zs header-read cols=+scoresums,+denom,-fid
  --score-col-nums 3-803
  --seed 31
  --threads 60

Start time: Thu Jan 12 11:55:05 2023
241801 MiB RAM detected; reserving 204800 MiB for main workspace.
Using up to 60 threads (change this with --threads).
399 samples (0 females, 0 males, 399 ambiguous; 399 founders) loaded from
vcf_Population_Report_ALL.psam.
57786076 variants loaded from vcf_Population_Report_ALL.pvar.zst.
Note: No phenotype data present.
Calculating allele frequencies... done.

Error: --score variant ID '2:168745247:A:AT' appears multiple times in main
dataset.
End time: Thu Jan 12 11:55:29 2023

[nebfield]

Hello,

Thanks for reporting the issue. It looks like when your VCF is converted into plink2 format, multiple variants have the same ID. This will cause a plink error when calculating a score.

One way to fix this is to edit a configuration file conf/modules.config. Change the following block:

    withName: PLINK2_VCF {
        ext.args = "--new-id-max-allele-len 100 missing"
    }

To:

    withName: PLINK2_VCF {
        ext.args = "--new-id-max-allele-len 100 missing --rm-dup force-first"
    }

This will deduplicate your variants when your VCF is recoded as a plink2 file. This should fix your error but you will lose some variants from the original VCF. You might want to check the plink documentation for other strategies.

plink2 -help  --rm-dup

Please let me know if you have any problems or questions.

[asmaa-a-abdelwahab]

@nebfield Thank you. The error is solved but I got another error:
Error: VCF file has a variant with 257 ALT alleles; this build of plink2 is limited to 254

[nebfield]

That's an error I've never seen before 😅 Some extra plink parameters may help:

withName: PLINK2_VCF {
    ext.args = "--new-id-max-allele-len 100 missing --rm-dup force-first --max-alleles 254"
}

I picked 254 based on your error message. If you want to remove multi-allelic variants you can set --max-alleles 2.

[asmaa-a-abdelwahab]

@nebfield Thanks for your reply. I already tried that. The executed command by default has --max-alleles 2 when I add --max-alleles 254 as an extra parameter, I get an error that the parameter is duplicated

Command executed:

  plink2 \
      --threads 60 \
      --memory 204800 \
      --set-all-var-ids '@:#:$r:$a' \
      --max-alleles 2 \
      --new-id-max-allele-len 100 truncate --rm-dup force-first \
      --vcf Merged_Population.vcf.gz  \
      --make-pgen vzs\
      --out vcf_Population_Report_ALL # 'vcf_' prefix is important
  
  cat <<-END_VERSIONS > versions.yml
  "PGSCATALOG_PGSCALC:PGSCALC:MAKE_COMPATIBLE:PLINK2_VCF":
      plink2: $(plink2 --version 2>&1 | sed 's/^PLINK v//; s/ 64.*$//' )
  END_VERSIONS

The error:

Command error:
  Error: Duplicate --max-alleles flag.

[nebfield]

Sorry, I forgot that flag was already included. It might be best to filter very complex multiallelic variants from your VCF using bcftools and using the filtered VCF as input to pgsc_calc, because plink is having trouble recoding your VCF into plink2 pfile format.

[asmaa-a-abdelwahab]

@nebfield Hello,
Sorry for the late reply and thanks for your recommendation, it did work for me