This function estimates the phenotypic effect of each accession's allele at
a particular marker. It is based on the probabilistic assignment of each
MAGIC line to a single accession. The procedure is repeated several times to
produce an average estimate and associated uncertainty. This function is based
on the function imputed.one.way.anova()
from the scripts available at
http://mtweb.cs.ucl.ac.uk/mus/www/magic/
estimateFounderEffect(x, phenotype, marker, n_samples = 500, summarised = TRUE, standardised = TRUE, covariates = NULL) # S4 method for MagicGenPhen estimateFounderEffect(x, phenotype, marker, n_samples = 500, summarised = TRUE, standardised = TRUE, covariates = NULL)
x | an object of class MagicData. |
---|---|
phenotype | the phenotype to get the results from. |
marker | the SNP marker to estimate the effects for. |
n_samples | number of Monte Carlo samples to draw. |
summarised | whether or not to report summarised results. The summarised output gives the mean, median and 2.5 imputations. The two percentiles define the 95 can be used as a confidence interval to report accuracy of the estimates. Default: TRUE |
standardised | whether or not the phenotypic values should be standardised to the mean. In that case the result reflects how many standard deviations the estimated effect deviates from the observed trait mean. Default: TRUE |
covariates | optional name of column(s) from phenotypes to use as covariate(s). The effect will be estimated from the residuals of a standard linear model between the trait of interest and the specified covariates. Default: NULL |
a data.frame with imputed phenotypes for each founder accession.
# NOT RUN { estimateFounderEffect(magic_genotypes, "MN1_29291") # }