Half-missing call in .ped file at variant 55, line 640.

[Shafagh Valipour]

Dear Dr. Chang,

I am trying to create .bed file from .ped and .map using this line of code:

plink --file "/scratch/.../Variants/20220622_Plink" --no-fid --no-parents --no-sex --no-pheno --aec --mind 0.1 --maf 0.05 --geno 0.1 --output-missing-genotype 0 --make-bed --out test

but, I am still getting an error that says : Half-missing call in .ped file at variant 55, line 640.

I checked line 55 of .ped file and there is "0 0" at the first column. I was wondering is there any way that I can solve this issue?

I appreciate your time and consideration.

Best Regards,
S

[Christopher Chang]

Note that the error message refers to line 640, not line 55.  The problem is with columns 110 and 111 of that line: with "--no-fid --no-parents --no-sex --no-pheno", the .ped should contain only one header column, then the 1st variant is in columns 2-3, the 2nd variant is in columns 4-5, ..., and the 55th variant is in columns 110-111.

Do you know how this .ped+.map file pair was generated?

[Shafagh Valipour]

Dear Dr. Chang,

Thank you so much for your reply. 

I received these files from genotyping company in plink ped+map format. 

Best Regards,

Shafagh

[Christopher Chang]

My first guess is that your "--no-fid --no-parents --no-sex --no-pheno" import flags aren't quite correct, then.  What are the first six columns of the first two lines?

[Shafagh Valipour]

This is how my .ped file looks like:

#Sample Filename    Genotypes
1000_208804_B    G G    T G    C C    G G    A G    T T    T T    T C    C C    A A    A G

I tried the command line without flags "--no-fid --no-parents --no-sex --no-pheno" too but that did not make any difference. I used this because this file does not have all 6 columns related to these flags. 

I would appreciate any suggestion that can fix this issue. 

Best Regards,

Shafagh

[Christopher Chang]

See my first response concerning line 640, columns 110-111 then.  What's there?

[Shafagh Valipour]

Here is the genotype at line 640:

A M4320_208959_H22.CEL    A G    T T    C C    G G    G G    T T    T T    T C    C C    A A    G G    T T    C C    T T    A G    T T    A A    G G    A A    T T    C C    T C    C C    A A    G G    G G    G G    A A    A A    A A    T C    C G    C C    G G    C C    G G    G G    C C    C C    C C    T C    G G    A A    A A    C C    C C    G G    C C    C C    G G    A A    A G    A A    G G    0 0    G G    G G    G G    T T    G G    C C    G G    C C    A A    C C    A A    A A    G G    G G    G G    T T    G G    T T    A A    T T    G G    C C    T T    T T    T T    A T    G G    G G    T T    A A    T T    T T    G G    C C    A A    G G    A A    T T    C C    G G    C C    C C    T T    G G    G G    T T    A A    T T    C C    G G    C C    A A    G G    A A    A A    A G    C C    T C

I appreciate your time. 

[Christopher Chang]

Well, it looks like you have a malformed .ped file.  This has two leading columns (--no-fid should be off), while an earlier line had one leading column.

If what you report is correct, this is really a failure of the genotyping company, but in the meantime you can use the data by standardizing the format to either 1-leading-column (then use --no-fid during import) or 2-leading-columns (don't use --no-fid).

[Shafagh Valipour]

So, it means an extra space in the sample name might caused this problem?

( A M4320_208959_H22.CEL instead of A_M4320_208959_H22.CEL).

[Christopher Chang]

Yes.

[elena dehnavi]

Hello Dr. Chang

 

I faced the same issue with plink 1.9. here is my variant

515707 MB RAM detected; reserving 257853 MB for main workspace.

Scanning .ped file...
Possibly irregular .ped line.  Restarting scan, assuming multichar alleles.

Error: Half-missing call in .ped file at variant 9, line 1.

 

I created ped file using python and check for any more space as well.

 

head -2 myfile.ped | cut -f7-20

0       0       2       2       2       1       2       2       0       0       0       0       0       0

1       2       2       2       2       1       2       1       1       0       1       0       0       0

 

kind regards,

Elena

chrch...@gmail.com در تاریخ دوشنبه ۸ اوت ۲۰۲۲ ساعت ۵:۴۵:۱۳ (UTC+10) نوشت:

[Christopher Chang]

How was the .ped file generated?  This does not satisfy the specification, which mandates two columns per variant, containing allele codes (e.g. "G G", or "0 0" representing a missing genotype).

You need to regenerate or transform this file into a format supported by plink.  If you have a choice, VCF is very strongly preferred over .ped nowadays.

[itziar lamiquiz moneo]

Dear Dr. Chang, 

I have the same problem. Ped and map file was generated by genotyping company. I have tried to convert to .bim, .bed and .fam files. However, I have this problem: 

Error: Half-missing call in .ped file at variant 1, line 2386.

Could you help me please?

Best regards, 

Itziar