SNP Genotype Data

This download includes two zip files and one text file

• GRLS_Genotyping_Axiom_Canine_Array_Sets_A_and_B_2023_08_02.zip
  • SetA
    • AxiomGT1.bin.bim
    • AxiomGT1.bin.fam
    • AxiomGT1.bin.bed
  • SetB
    • AxiomGT1.bin.bim
    • AxiomGT1.bin.fam
    • AxiomGT1.bin.bed
  • map_id_sex.tab*
• GRLS_Genotyping_Axiom_Canine_Merged_2023_08_02.zip
  • AxiomGT1v2.merged.bim
  • AxiomGT1v2.merged.fam
  • AxiomGT1v2.merged.bed
• README.txt

*The map_id_sex maps between replicate IDs in the genotyping files and their corresponding subject IDs in the phenotypic data. The family ID refers to the three different sample groups: 

• PILOT - a small subset of the GRLS cohort that were used in a proof of concept for the genotyping project
• GRLS - the reminder of the GRLS cohort
• Oldies - from the Golden Oldies subproject

GRLS Genotyping Axiom Canine Array Sets A and B

~1.1 million genetic markers from the entire Golden Retriever Lifetime Study (GRLS) cohort and 200 of the Golden Oldie’s* participants were genotyped using Affymetrix (ThermoFisher) Axiom™ Canine Genotyping Array Sets A and B. 

This zip file contains the raw data from both Array A and B that must be merged into a final dataset. There are some issues with the data that must be resolved when merging:

• Some dogs have multiple replicates in the output files. These helped with quality control during the processing but only one will be needed when using the data. 
  • Example: dog grls9ZOSMBXX has multiple replicates with the IDs grls9ZOSMBXX_1, grls9ZOSMBXX_2 and grls9ZOSMBXX_3.
• Two samples have an extremely high rate of mismatching genotypes between the two array sets -- grlsU85RZ2FF_1 and grlsQZAMFHKK_1.
• 1463 markers that have the same position on both arrays but have different SNP IDs. The array annotation shows that most of these markers are identical with minor differences in the flanking sequences.
• There are two duplicates of a sample that, after looking at the KING-robust kinship estimator, are from different dogs -- grlsFGFIVFXX_1 and grlsFGFIVFXX_2.
• There are some dogs with different genders on both arrays and within the phenotypic data:
  • Four are listed as female in the PLINK files and male in the phenotype files -- grlsI0RDZYGG, grlsKLCB9Q55, grlsMJNZJKDD and grlsZJRGT144.
  • Five are listed as male in the PLINK files and female in the phenotype files -- grlsYOQLXM44, grls7YEZWGBB, grls4THCKBII, grlsYS5D3CVV and grlsYY0S6W44.
  • One is listed as female on Array A and male on Array B -- grlsQZAMFHKK_1.
  • One is listed as male in the PLINK file and phenotype files, but has heterozygous genotypes on the X chromosome while missing the Y-specific markers -- grlsJCVWWRMM.
• There are 18 dogs that are missing >40% of their genotypes.
• Further quality control by the individual may result in more samples being excluded from analysis.

A step-by-step tutorial for how to perform a GWAS study utilizing the GRLS phenotypic data and for how to address the known issues with the genetic data as outlined above has been created by Dr. Tamer Mansour, an Assistant Research Scientist at the University of California, Davis. That tutorial can be found here.

GRLS Genotyping Axiom Canine Merged

~1.1 million genetic markers from the entire Golden Retriever Lifetime Study (GRLS) cohort and 200 of the Golden Oldie’s* participants were genotyped using Affymetrix (ThermoFisher) Axiom™ Canine Genotyping Array Sets A and B. 

This zip file contains a copy of the genetic data that has had both array data merged and has had each of the previously outlined issues (see GRLS Genotyping Axiom Canine Array Sets A and B above for the list of issues) addressed. For dogs with multiple replicates, any replicate with a low genotyping rate or quality were dropped. For those that still had more than one replicate, a replicate was chosen at random to be included in the final output. For more information about the processing that was done, please refer to the Prep Array Sets section of the GRLS GWAS Tutorial by Dr. Tamer Mansour.

*A subproject of the Golden Retriever Lifetime Study was The Golden Oldies study. In this project, an additional cohort of golden retrievers that had reached the age of 12 years or older and had not yet been diagnosed with cancer were recruited to provide DNA samples. We anticipated this population could serve as a control group. Only DNA samples were collected from this group of dogs. Because the purpose of this study was to create an additional sample inventory, there is no phenotypic data from questionnaires available.

The DOI for this dataset is https://doi.org/10.62291/ACMJ1341