Two-Phase Genetic Association Study Design and Analysis with Missing Covariates by Design R package twoPhaseGAS version 1.2.5 Provides functionality for designing and analysing two-phase genetic association studies. Phase 1 data usually come from genome-wide association study (GWAS) results and we assume phase 2 data will be part of a targeted genome sequencing or fine-mapping study. At design stage, the package assists in selecting a subset of individuals that will be sequenced for phase 2 via alternative approaches, including a flexible genetic algorithm (GA) for near-optimal designs. Once phase 2 data have been collected, the package implements methods to analyse phase 1 and phase 2 data together using semi-parametric regression models via the expectation-maximization (EM) algorithm. For more details see Espin-Garcia, Craiu and Bull (2018) <doi:10.1002/gepi.22099> and Espin-Garcia, Craiu and Bull (2021) <doi:10.1002/sim.9211>. Software Depends: R (>= 3.5.0) Imports: data.table, dfoptim, enrichwith, kofnGA (>= 1.2), MASS, Matrix, nloptr, parallel, stats, utils Suggests: gplots, knitr, rmarkdown, testthat (>= 3.0.0) Osvaldo Espin-Garcia <oespinga@uwo.ca> Comprehensive R Archive Network (CRAN) Osvaldo Espin-Garcia [aut, cre], Apostolos Dimitromanolakis [aut], Radu V Craiu [ths], Shelley B Bull [ths] GPL (>= 2) 2026-05-27 application/tgz https://CRAN.R-project.org/package=twoPhaseGAS doi:10.32614/CRAN.package.twoPhaseGAS