# Wrong format .ped in PLINK

I know that I don't have the correct format in my .ped file for PLINK but I don't know how to convert it. My actual file is something like:

BSCAN BSCAN20016 0 0 2 0 0 1 1 1....
BSCAN BSCAN20017 0 0 2 0 0 1 1 1....
BSCAN BSCAN20018 0 0 2 0 0 1 1 1....


Where column 1 is FamilyID, column 2 is ID, column 3 is FatherID, column 4 MotherID, column 5 Sex and then the genotype of each individual. I only need to create binary files, so I tried to run:

plink-1.9-rc --file admixed --make-bed --no-pheno --cow --out admixed


but the program shows: Error: half-missing call in .ped file at variant 1, line 1.

Anyone know how can I change the format or recode it to make a correct input?

So, columns 7 and 8 should be the genotype of variant 1 (each column is one of the alleles). You have them as 0 1, which means absent and major allele.

I think you're just missing a column. From the doc:

Column 6 is Phenotype value ('1' = control, '2' = case, '-9'/'0'/non-numeric = missing data if case/control)

Ideally you would know which SNPs are heterozygous based on your SNP calling and have such SNPs as 1 2, but if you have a homozygous organism or if you're working with data that's been flattened you can simply generate the .ped file with the same number twice for each SNP.

• Thanks for your reply. So, I need to phase the data? Because now I have one column per SNP. I don't have the phenotype column but I specify in the command "--no-pheno" – aguillenea Jan 17 '20 at 14:47
• Yeah you're right that --no-pheno should mean it can read them fine, but it says specifically that the first variant in the first column is half-empty. Can you try after adding a phenotype column? You can use -9 to indicate you don't have one. – Thymine Jan 17 '20 at 15:02
• Oh, is your example not your real data maybe? I assumed it was real aside from there being more lines and columns – Thymine Jan 17 '20 at 15:04
• Actually, it is quite similar to my real data... but I have 5564 individuals (rows) and 46465 SNPs (columns). I will try adding the phenotype column... – aguillenea Jan 17 '20 at 15:07
• I have the same error with the phenotype column. – aguillenea Jan 17 '20 at 15:15