Genome-wide association study (GWAS) and genomic prediction/selection (GP/GS) are the two essential enterprises in genomic research. Due to the great magnitude and complexity of genomic and phenotypic data, analytical methods and their associated software packages are frequently advanced. GAPIT is a widely-used genomic association and prediction integrated tool as an R package. The first version was released to the public in 2012 with the implementation of the general linear model (GLM), mixed linear model (MLM), compressed MLM (CMLM), and genomic best linear unbiased prediction (gBLUP). The second version was released in 2016 with several new implementations, including enriched CMLM (ECMLM) and settlement of MLMs under progressively exclusive relationship (SUPER). All the GWAS methods are based on the single-locus test. For the first time, in the current release of GAPIT, version 3 implemented three multi-locus test methods, including multiple loci mixed model (MLMM), fixed and random model circulating probability unification (FarmCPU), and Bayesian-information and linkage-disequilibrium iteratively nested keyway (BLINK). Additionally, two GP/GS methods were implemented based on CMLM (named compressed BLUP; cBLUP) and SUPER (named SUPER BLUP; sBLUP). These new implementations not only boost statistical power for GWAS and prediction accuracy for GP/GS, but also improve computing speed and increase the capacity to analyze big genomic data. Here, we document the current upgrade of GAPIT by describing the selection of the recently developed methods, their implementations, and potential impact. All documents, including source code, user manual, demo data, and tutorials, are freely available at the GAPIT website (http://zzlab.net/GAPIT).

Genome-Wide Association Study (GWAS) and Genomic Prediction/Selection (GP/GS) are the two essential enterprises in genomic research. Due to the great magnitude and complexity of genomic data, analytical methods and their associated software packages are frequently advanced. GAPIT is a widely used Genomic Association and Prediction Integrated Tool. The first version was released to the public in 2012 with the implementation of the general linear model (GLM), mixed linear model (MLM), compressed MLM, and genomic Best Linear Unbiased Prediction (gBLUP). The second version was released in 2016 with several new implementations, including Enriched Compressed MLM and Settlement of mixed linear models Under Progressively Exclusive Relationship (SUPER). All the GWAS methods are based on the single locus test. For the first time, in the current release of GAPIT, version 3 implemented three multiple loci test methods, including Multiple Loci Mixed Model (MLMM), Fixed and random model Circulating Probability Unification (FarmCPU), and Bayesian-information and Linkage-disequilibrium Iteratively Nested Keyway (BLINK). Additionally, two GP/GS methods were implemented based on Compressed MLM, named compressed BLUP, and SUPER, named SUPER BLUP. These new implementations not only boost statistical power for GWAS and prediction accuracy for GP/GS, but also improve computing speed and increase the capacity to analyze big genomic data. Here, we document the current upgrade of GAPIT by describing the selection of the recently developed methods, their implementation, and potential impact. All documents, including source code, user manual, demo data, and tutorials, are freely available at the GAPIT website (http://zzlab.net/GAPIT).

Improvement of statistical methods is crucial for realizing the potential of increasingly dense genetic markers. Bayesian methods treat all markers as random effects, exhibit an advantage on dense markers, and offer the flexibility of using different priors. In contrast, genomic best linear unbiased prediction (gBLUP) is superior in computing speed, but only superior in prediction accuracy for extremely complex traits. Currently, the existing variety in the BLUP method is insufficient for adapting to new sequencing technologies and traits with different genetic architectures. In this study, we found two ways to change the kinship derivation in the BLUP method that improve prediction accuracy while maintaining the computational advantage. First, using the settlement under progressively exclusive relationship (SUPER) algorithm, we substituted all available markers with estimated quantitative trait nucleotides (QTNs) to derive kinship. Second, we compressed individuals into groups based on kinship, and then used the groups as random effects instead of individuals. The two methods were named as SUPER BLUP (sBLUP) and compressed BLUP (cBLUP). Analyses on both simulated and real data demonstrated that these two methods offer flexibility for evaluating a variety of traits, covering a broadened realm of genetic architectures. For traits controlled by small numbers of genes, sBLUP outperforms Bayesian LASSO (least absolute shrinkage and selection operator). For traits with low heritability, cBLUP outperforms both gBLUP and Bayesian LASSO methods. We implemented these new BLUP alphabet series methods in an R package, Genome Association and Prediction Integrated Tool (GAPIT), available at http://zzlab.net/GAPIT.