Rapid and Accurate Genetic Prediction Modeling for Genome-Wide Association or Whole-Genome Sequencing Study Data

Rapidly build accurate genetic prediction models for genome-wide association or whole-genome sequencing study data by smooth-threshold multivariate genetic prediction (STMGP) method. Variable selection is performed using marginal association test p-values with an optimal p-value cutoff selected by Cp-type criterion. Quantitative and binary traits are modeled respectively via linear and logistic regression models. A function that works through PLINK software (Purcell et al. 2007 , Chang et al. 2015 ) < https://www.cog-genomics.org/plink2> is provided. Covariates can be included in regression model.

Overview

Include functions to rapidly build accurate genetic prediction models for genome-wide association or whole-genome sequencing study data by smooth-threshold multivariate genetic prediction (STMGP) method.

Installation

To install: install.packages("stmgp")

References

Ueki M, Tamiya G, and for Alzheimer's Disease Neuroimaging Initiative. (2016) Smooth-thresholdmultivariate genetic prediction with unbiased model selection. Genet Epidemiol 40:233-43. https://doi.org/10.1002/gepi.21958 Ueki M. (2009) A note on automatic variable selection using smooth-threshold estimating equations. Biometrika 96:1005-11. https://doi.org/10.1093/biomet/asp060