mxr_associate: Associate the genotype with the phenotype.

Description Usage Arguments Details Value


mxr_associate outputs the results of the mixed linear model association between the SNPs and the phenotype. It also outputs the REsidual Maximun Likelihood (REML) estimates as well as the pseudoheritability estimate. All calculations are done using EMMAX


mxr_associate(tped_prefix, pheno_file, kin_file, out_prefix, cov_file = "",
  verbose = F)



The path to the transposed '12' recoded plink file. The A1 allele (the minor allele – or effect allele) is recoded as "1", while A2 allele is recoded as "2". The complete call to plink is 'plink2 --bfile [bed_prefix] --recode12 --output-missing-genotype 0 --transpose --out [tped_prefix]'.


The path to the phenotype file. There are three required columns: FID (family ID), IID (individual ID), and phenotype column. It is recommended that the phenotype column be approximately normally distributed, in which case, the warpedLMM software would be very useful. IMPORTANT: The samples need to be sorted according to how they are arranged in the genotype file. Missing phenotype values must be represented as 'NA'.


The path to the kinship matrix file computed using emmax-kin. Use the Balding-Nichols algorithm. Generate the kinship matrix by a call to emmax-kin 'emmax-kin-intel64 -v -d 10 [tped_prefix]'.


Path and prefix of the output files.


(Optional) Path to the covariate file. The first two columns of the covariate file must be the FID and IID of the samples. The third column must be a vector of '1'. Quantitative covariates start from the fourth column onward.


(Optional) Show verbose output. (DEFAULT=FALSE)


Before running this function, make sure to compute EMMAX to compute the association of each SNP, coded as minor allele additive dosage (using –recode A in PLINK), with the phenotype.


TRUE if the EMMAX run completed successfully. FALSE, otherwise.

roslen/mxr documentation built on May 27, 2019, 11:32 p.m.