Description Usage Arguments Details Value Note Author(s) See Also Examples

An internal, function for fast association testing. It tests the phenotype of interest for association with methylation coverage (columns of the data parameter).

1 | ```
testPhenotype(phenotype, data1, cvrtqr)
``` |

`phenotype` |
Vector with phenotype. Can be numerical, character, or factor vector. |

`data1` |
Matrix with data (normalized coverage), one variable (CpG) per column. |

`cvrtqr` |
Orthonormalized covariates, one covariate per column.
See |

The testing is performed using matrix operations and C/C++ code, emplying an approach similar to that in MatrixEQTL.

If the phenotype is numerical, the output is a list with

`correlation` |
Correlations between residualized phenotype and data columns. |

`tstat` |
Corresponding T-statistics |

`pvalue` |
Corresponding P-values |

`nVarTested` |
Always 1 |

`dfFull` |
Number of degrees of freedom of the T-test |

If the phenotype is a factor (or character)

`Rsquared` |
R-squared for the residualized ANOVA F-test. |

`Fstat` |
Corresponding F-test |

`pvalue` |
Corresponding P-values |

`nVarTested` |
First number of degrees of freedom for the F-test. Equal to the number of factor levels reduced by 1 |

`dfFull` |
Second number of degrees of freedom for the F-test. |

This function is used in several parts of the pipeline.

Andrey A Shabalin [email protected]

See vignettes: `browseVignettes("ramwas")`

.

Also check `orthonormalizeCovariates`

.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 | ```
### Generate data inputs
# Random data matrix with signal in the first column
data = matrix(runif(30*5), nrow = 30, ncol = 5)
data[,1] = data[,1] + rep(0:2, each = 10)
# Two random covariates
cvrt = matrix(runif(2*30), nrow = 30, ncol = 2)
cvrtqr = orthonormalizeCovariates(cvrt)
### First, illustrate with numerical phenotype
# Numerical, 3 value phenotype
phenotype = rep(1:3, each = 10)
# Test for association
output = testPhenotype(phenotype, data, cvrtqr)
# Show the results
print(output)
# Comparing with standard R code for the first variable
summary(lm( data[,1] ~ phenotype + cvrt ))
### First, illustrate with numerical phenotype
# Categorical, 3 group phenotype
phenotype = rep(c("Normal", "Sick", "Dead"), each = 10)
# Test for association
output = testPhenotype(phenotype, data, cvrtqr)
# Show the results
print(output)
# Comparing with standard R code for the first variable
anova(lm( data[,1] ~ cvrt + phenotype ))
``` |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.