In the examples, we will go over how the package LTFHPlus can be used to estimate a refined phenotype with the LT-FH++ package. More specifically, we will provide reproducible examples with explanations of the method and its code, such that users will be able to play around with the functions themselves. The code will be based on packages from the tidyverse, if the reader is not familiar with packages such as dplyr, then a line-by-line understanding might not be possible, however the overall point should hopefully still be clear.

A word on the examples

The examples included here are meant to represent a simple application of the method, such that users are able to get a good idea of how to use the functions in the package and what to expect as output. The genetic liabilities are meant to be used as a phenotype for genome-wide association studies(GWAS). We recommend using software such as PLINK, BOLT, Regenie, or R-based methods such as bigsnpr to perform the GWAS.

What would you like to read about?

Here is an overview of what you can learn more about:

  • From CIP and family pedigree to input
    • What does the CIP and family information look like?
    • What the input format is like
  • How the covariance matrix is constructed
    • What format should the family relationship be like
    • How does the covariance matrix use the family relationship
    • Example
  • LT-FH++ Example
    • Simulate phenotypic data under the liability threshold
    • Estimating genetic liability with parallelization