Reference allel error in ped to vcf conversion

Question

[mel22]

Hi, In view to phase my data, i converted the ped file to vcf with plink, but the problem I have continuously a problem concerning reference alleles what is the best way to conver ped/bed file to vcf and align it to the reference genome to avoid the problem with refernce allele? thank you

[Kevin Blighe]

Please provide more information. What is the source of your data? How was it processed? Which program(s) are you using?

[mel22]

Hi Kevin, My data come from genotyping with illumina chip, I tried to phase with eagle and shapeit but you have the same problem in both programs. For the vcf conversion I used this :

plink --file my file  --recode vcf --keep allele order

Thanks

[Kevin Blighe]

What are the exact error and warning messages that you are receiving? Indeed, when using data from genotyping microarrays, not all SNPs will be aligned to the reference. You may have to obtain the library annotation for your array type in order to identify these, and/or make use of the --flip parameter in PLINK.

[mel22]

Yes how to get the vcf file from plink with the same allele order of the library oth array ? For example, for phasing from CHR1 , I have more than 30000 SNP excluded because of the ref/alt alleles difference between data and reference ?

Thanks

[Kevin Blighe]

"What are the exact error and warning messages that you are receiving?"

[mel22]

Example for chr 1 output from shapeit:

  Reference and Main panels are not well aligned:
      * #Missing sites in reference panel = 33668
      * #Misaligned sites between panels = 707
      * #Multiple alignments between panels = 0

[Kevin Blighe]

Please be aware that you need to run SHAPEIT like this: C: Phasing with SHAPEIT

What is your reference panel, by the way?

[mel22]

I used 1000 G as reference panel , the problem that If I will exclude problematic SNPs I will have to exclude about 30000 SNPs in chr1 for example ...