Plink genotype data (based on SNP) format!

Question

[kyc92dream]

Hi there,

My genotype data only has 2 types and designed as 0 and 1 respectively (they are not in pairs format only give single number, and the third is ? represents missing), and i have no idea how could i arrange my ped file format as it requires 2 columns for each SNP. I tried when i set one column for each SNP, it indicated that the found columns less than expected. Is there anyone who can help me with this so i could arrange my genotype data to fit running format?

[GabrielMontenegro]

Could you provide an example of your data?

[kyc92dream]

0 H1 0 0 0 0.9 1 0 0 0 1 1 1 1 0

0 H2 0 0 0 2.3 0 0 0 0 1 1 1 0 0

0 H3 0 0 0 1.1 0 0 1 0 1 1 1 0 0

0 H4 0 0 0 1.1 0 0 0 0 1 1 1 0 0

0 H5 0 0 0 0.4 0 0 0 0 1 1 1 0 0

0 H6 0 0 0 1.1 0 0 0 0 1 0 1 0 0

I am doing quantitive trait study in crop, no family ID, parental ID and sex where i typed 0 under corresponding columns. SNP-based genotype data starts from 7th column(i only pasted 6 individuals and 9 SNPs here), one SNP takes one column.

[kyc92dream]

And i tried to add flag compound-genotype, but it did not work as it should be 2 characters long.

Answer 1

[MoppelKopp]

You need an additional Map file. With that, plink should be able to convert your files automatically.

Example from Plink's own "toy" data:

(without header) CHR SNP-ID cM_Position(set to 0) BP_position
1 rs0 0 1000
1 rs10 0 1001

just create this file with a line for each of your markers and without the header, and it should work.

plink command example:
"plink --file INPUT --make-bed --out OUTFILE"

Plink will create a "binary" set of three files:

• .fam - phenotype information for each sample
• .bed - binary genotype file
• .bim - map file

The map file will contain additional columns for your genotype alleles, coded with 0/1/2. You may later change that if you like, see "https://www.cog-genomics.org/plink/1.9/data#update_map"

Some additional tips:

• If you don't have a family ID, use the sample ID for the family ID as well. For now, al your samples belong to family "0"
• Use Plink 2.0, it is much faster. Keep Plink 1.07 as a backup copy somewhere, some features of Plink 1.07 are not implemented in Plink 2 yet

[MoppelKopp]

I forgot something: You have to change the missing identifier (plink standard is "-9") or add the option "--missing-code ?" - but I am not sure, if plink is able to process the "?".