Changes in version 2.12.0 o fix a C++ error that 'span' is ambiguous o new functions `seqRefitNullGLMM()`, `seqAssocGLMM_GT()` and `seqAssocGLMM_ACAT_O_GT()` o SAIGEgds supports optional OpenCL-based GPU acceleration for the GRM cross-product computation during null model fitting o new `use.gpu` argument in `seqFitSparseGRM()` and `seqFitDenseGRM()` for OpenCL GPU-accelerated GRM construction o new 'geno.model' argument in `seqAssocGLMM_SPA()`, `seqAssocGLMM_GT()`, `seqAssocGLMM_Burden()`, `seqAssocGLMM_SKAT()`, `seqAssocGLMM_ACAT_V()`, `seqAssocGLMM_ACAT_O()` and `seqAssocGLMM_ACAT_O_GT()` for additive, dominant, or recessive genetic models Changes in version 2.10.2 o 'fork.loading' is deprecated in `seqFitNullGLMM_SPA()`, use 'parallel.loading' instead Changes in version 2.9.2 o improve the work load balancing in `seqAssocGLMM_SPA()` according to SeqArray_1.49.6 o new option 'verbose.pval' to improve the display in `seqAssocGLMM_SPA()` Changes in version 2.4.1 o add an option `load.balancing` to `seqAssocGLMM_SPA()` Changes in version 2.4.0 o update `seqFitLDpruning()` o reduce the memory usage in `seqAssocGLMM_SPA()` when the genotype file is not split by chromosomes Changes in version 2.2.1 o fix the compiler warning: -Wformat-security o fix a compiler issue with gcc v7.5 Changes in version 2.2.0 o fix the "Matrix-deprecated" warning from the Matrix package when calling `as(, "dgCMatrix")` Changes in version 2.0.0 NEW FEATURES o New features to use sparse genetic relationship matrix in generalized linear mixed models (GLMMs) according to the SAIGE-GENE paper (Zhou et al., 2020) o a new argument 'grm.mat' in `seqFitNullGLMM_SPA()`: the dense or sparse genetic relationship matrix (GRM) can be specified via 'grm.mat' o MAC categories for multiple variance ratios in `seqFitNullGLMM_SPA()`, `seqAssocGLMM_SPA()`, designed for rare variants o new function `seqAssocGLMM_SKAT()` for the SKAT aggregate method o new feature `seqAssocGLMM_ACAT_O()` to include the SKAT aggregate method for full ACAT-O tests o faster `seqSAIGE_LoadPval()` when merging multiple input files Changes in version 1.12.5 o enable RCPP_PARALLEL_USE_TBB=1 on Windows Changes in version 1.12.4 o fix the compiling issue on ARM64 (see https://github.com/AbbVie-ComputationalGenomics/SAIGEgds/issues/8) Changes in version 1.12.1 o fix the memory issue because of using deprecated tbb::task_scheduler_init in RcppParallel Changes in version 1.6.0 o `seqFitNullGLMM_SPA()` can use imputed dosages directly without converting the dosages to the best-guess genotypes o new function `glmmHeritability()` for approximate heritability estimates Changes in version 1.4.0 o set-based tests: burden, ACAT-V Changes in version 1.2.2 o update the citation o work around gcc-10 Changes in version 1.2.0 o `seqAssocGLMM_SPA()` can save to a RDS file Changes in version 1.0.2 o fix an issue when there is no covariate in the formula (e.g., y ~ 1) Changes in version 1.0.1 o add a new option for the threshold of missing variant to `seqAssocGLMM_SPA()` o change the output column 'AC.alt' to 'mac' in `seqAssocGLMM_SPA()` o retry model fitting at least 10 times if matrix is singular or not positive definite, or large variance is observed Changes in version 1.0.0 o first Bioconductor release Changes in version 0.99.0 o package submitted to Bioconductor Changes in version 0.9.10 o SAIGE algorithm implementation for quantitative outcomes Changes in version 0.9.9 o add a vignette to the package o the default of random number generator changes in R: "Rounding" was the default in `RNGkind()` prior to R_3.6.0, but "Rejection" is used in R (>= v3.6.0). For reproduction of the results created by R (< v3.6.0), please use `RNGkind("Mersenne-Twister", "Inversion", "Rounding")` in R (>= v3.6.0) Changes in version 0.9.7 o seqAssocGLMM_SPA(): load balancing in parallel Changes in version 0.9.0 o first version of SAIGEgds