ShapeIT check problematic SNPs

Question

[SHN]

Hello All,

I am using shapeit to phase my genotype data using 1000G phase3 as a reference genome. In order to phase my genotypes, I am using SHAPEIT. in the first step, I am using the command shapeIT to remove the problematic SNPs as has been mentioned in the link below: Phasing with SHAPEIT

However, when in ran the second round of SHAPEIT -check to exclude the problematic variants using --exclude-snp Prephased/MyData_chr"${chr}"_alignments.snp.strand.exclude (the link above), I still get error and files with alignment.snp.strand.exclude.

So here are my questions:

• Is it correct not to get all the problematic SNPs/variants in the first round of shapeit --check?
• Should I continue this step (below) again to remove the newly found problematic SNPs/variants until I get no error?

Thank you for your responses,

for chr in X {1..22}; do 

  plink --bfile MyData --chr "${chr}" --make-bed --out temp

  if [ "${chr}" != "X" ]

  then

    srun --mem=8 --cpus-per-task=4 --partition=serial \

      shapeit \

        -check \

        -B temp \

        -M library/1000GP_Phase3/genetic_map_chr"${chr}"_combined_b37.txt \

        --input-ref library/1000GP_Phase3/1000GP_Phase3_chr"${chr}".hap.gz library/1000GP_Phase3/1000GP_Phase3_chr"${chr}".legend.gz library/1000GP_Phase3/1000GP_Phase3.sample \

        --exclude-snp Prephased/MyData_chr"${chr}"_alignments.snp.strand.exclude \
        -T 8 ;
  fi

done ;

rm temp.* ;

[Kevin Blighe]

Hi, I wrote the code in the other threads. Can you show the first command for step 1? What is the error message(s) that you are receiving?

[SHN]

Hi Kevin,

Thanks for your message. I actually am doing the same steps. here is the code for the first step:

for i in {1..22};do
echo "* Chromosome: "$i
touch $data_path/checktemp
plink --bfile $start_file_name --chr "$i" --make-bed --bp-space 1 --out $result_unpahsed/temp  &> $data_path/checktemp && echo "  -- First step Run Checks!"
echo "*Reference haplotype: "$reference_data_path/1000GP_Phase3_chr"$i".hap.gz

if [ "$i" != "X" ]; then
    $SHAPEIT -check \
        -B $result_unpahsed/temp \
        -M $reference_data_path/genetic_map_chr"$i"_combined_b37.txt \
        --input-ref $reference_data_path/1000GP_Phase3.sample $reference_data_path/1000GP_Phase3_chr"$i".hap.gz $reference_data_path/1000GP_Phase3_chr"$i".legend.gz \
        --output-log $result_unpahsed/data_qced_chr"$i"_alignments \
        -T 8;
fi
done

After this step, I get files data_qced_chr2_alignments.snp.strand.exclude" and "data_qced_chr2_alignments.snp.strand".

I use this file "_alignments.snp.strand.exclude" for the second round of shapeit -checks and I still get missing sites with variants saved in ".snp.strand.exclude".

and the output files is as below:

MODE ESC[33m-summariseESC[0m : GENERATING SUMMARY STATISTICS OF THE INPUT DATA

• Autosome (chr1 ... chr22)
• Reference panel of haplotypes used

Parameters :

• Seed : 1620163254
• Parallelisation: 1 threads
• Ref allele is NOT aligned on the reference genome

Reading SNPs to exclude from input file in [path/data_qced_chr5_alignments.snp.strand.exclude]

• 96128 snps found in the exclude list

Reading site list in [PATH/qc_plink/unphased_chr/temp.bim]

• 21882 sites included
• 12 sites excluded

Reading sample list in [PATH/qc_plink/unphased_chr/temp.fam]

• 328 samples included
• 328 unrelateds / 0 duos / 0 trios in 328 different families

Reading genotypes in [/PATH/qc_plink/unphased_chr/temp.bed]

• Plink binary file SNP-major mode

Reading sample list [/PATH/1000GP_Phase3/1000GP_Phase3_chr5.legend.gz]

• 10587830 reference haplotypes included

Reading SNPs in [PATH/1000GP_Phase3/1000GP_Phase3_chr5.hap.gz]

• 0 reference panel sites included
• 5293914 reference panel sites excluded

ERROR: Reference and Main panels are not well aligned:

• #Missing sites in reference panel = 21882
• #Misaligned sites between panels = 0
• #Multiple alignments between panels = 0

Should I continue excluding variants with another round of shapeit -check and then prephasing? Or are these the new variants that should be used for the prephasing step (step3 in your codes) mentioned here: Phasing with SHAPEIT

Thanks,

[Kevin Blighe]

Hello again. hmmm, the --exclude-snp does not actually remove variants; so, if you run it again, it will just produce the same result.

I wonder have you done QC on your input data? Which array is it? Prior to doing the imputation, the code for which I shared here on Biostars in the other threads, I first performed QC, and then successfully pre-phased and imputed ImmunoChip and Illumina GSA again 1000 Genomes Phase III.

By QC, I mean filtering out variants with high missingness, multi-allelic records, duplicates, etc

---

---

You can also just try to proceed with the pre-phasing - not sure.

[SHN]

Thanks for your input. Yes I did the QC on the genotypes, ran shapeit -check, running step2! and I stopped here!

I thought that when we exclude variants in the second step, then the program should not produce additional variants.

My array is Infinium omni2.5, Illumina and I want to prophase and impute it against 1000G phase III.

un-alignments can affect imputation accuracy. I should probably redo everything just to make sure everything is running correctly.