I am currently performing a GWAS using a Illumina genotyping chip and PLINK2.
As part of the preprocessing I looked into the population stratification of my cohort and observed that all individuals come from Europe (which is what I also expected). Nonethless, I included the first 10 principal components as covariates in the linear association model. PLINK2 returned a genomic inflation estimate of 1.00085. If I rerun the analysis, but exclude the PC as covariates, I still get a genomic inflation estimate of 1.00159 which is still very close to 1.
So my question is, if I even need to include the PCs as covariates as it seems like there is no population stratification at all.
0 answers
No answers yet.
Log in to answer this question.
-
Functional normalization using minfi::preprocessFunnorm and control PCs in later analyses
written by mendopsy •
-
genomic inflation factor (lambda)=1
written by putty •
-
Adding Principal components as covariates in GWAS studies
written by anikcropscience •
-
GWAS principal components
written by Raygozak
-
PCA: summarize results from principal components analysis and look for population heterogeneity
written by F_cm_C •
-
Genomic inflation in genome-wide survival analysis
written by Caragh •
-
PLINK Principal Components not adequately controlling for population stratification in linear regre…
written by dam4l
-
Interpretation of Eigenstrat results
written by kelly.wang135
-
Gwas: Adjust For Population Stratification In Impute2 Output
written by w0nderingatwork •
-
Ok To Use Pcs Computed On Entire Sample For Gwas In Subset Of Sample?
written by Stephen
