Greetings! I am trying to perform a linear regression analysis with SNP genotypes as covariates because i would like to assess the associations between phenotypes and combinations of SNPs. I have 11 SNPs in my binary files, and a bunch of phenotypes in my .pheno file. If i perform linear regression without covariates, or with various phenotypes as covariates, i have no problems and get sensical results. However, when i try to perform linear regression with SNP genotypes as covariates, i get a funny error. As you can probably guess i did not find a solution through google.
Here is the info (highlighted the lines that i don't get):
Options in effect:
--all-pheno
--allow-no-sex
--bfile plink
--ci 0.95
--covar plink.cov
--covar-name rs1,rs2,rs3
--linear
--missing-phenotype 99999
--pheno plink_pheno.pheno
11 variants loaded from .bim file.
519 people (220 males, 286 females, 13 ambiguous) loaded from .fam.
Ambiguous sex IDs written to plink.nosex .
313 phenotype values present after --pheno.
Using 1 thread (no multithreaded calculations invoked).
--covar: 3 covariates loaded.
**519 people had missing value(s).**
Before main variant filters, 519 founders and 0 nonfounders present.
Calculating allele frequencies... done.
Total genotyping rate is 0.696269.
11 variants and 519 people pass filters and QC.
**Note: No phenotypes present.**
**Warning: Skipping --linear since # variables >= # samples.
(Check your covariates--all samples with at least one missing covariate are
excluded from this analysis.)**
Can you post the first 10 lines of the plink.fam and plink.cov files?
I'm also facing the same error. I just have 10 samples. What will be the reason for this error? could you please help me out?
Thanks, Karthick
As I noted at https://github.com/chrchang/plink-ng/issues/144 , you don't have enough samples. See e.g. https://stats.stackexchange.com/questions/26016/sample-size-for-logistic-regression .