Zhiwu Zhang

TASSEL: software for association mapping of complex traits in diverse samples

Association analyses that exploit the natural diversity of a genome to map at very high resolutions are becoming increasingly important. In most studies, however, researchers must contend with the confounding effects of both population and family structure. TASSEL (Trait Analysis by aSSociation, Evolution and Linkage) implements general linear model and mixed linear model approaches for controlling population and family structure. For result interpretation, the program allows for linkage disequilibrium statistics to be calculated and visualized graphically. Database browsing and data importation is facilitated by integrated middleware. Other features include analyzing insertions/deletions, calculating diversity statistics, integration of phenotypic and genotypic data, imputing missing data and calculating principal components.

Genome-wide association studies of 14 agronomic traits in rice landraces.

Uncovering the genetic basis of agronomic traits in crop landraces that have adapted to various agro-climatic conditions is important to world food security. Here we have identified ∼3.6 million SNPs by sequencing 517 rice landraces and constructed a high-density haplotype map of the rice genome using a novel data-imputation method. We performed genome-wide association studies (GWAS) for 14 agronomic traits in the population of Oryza sativa indica subspecies. The loci identified through GWAS explained ∼36% of the phenotypic variance, on average. The peak signals at six loci were tied closely to previously identified genes. This study provides a fundamental resource for rice genetics research and breeding, and demonstrates that an approach integrating second-generation genome sequencing and GWAS can be used as a powerful complementary strategy to classical biparental cross-mapping for dissecting complex traits in rice.

GAPIT: Genome Association and Prediction Integrated Tool

SUMMARY Software programs that conduct genome-wide association studies and genomic prediction and selection need to use methodologies that maximize statistical power, provide high prediction accuracy and run in a computationally efficient manner. We developed an R package called Genome Association and Prediction Integrated Tool (GAPIT) that implements advanced statistical methods including the compressed mixed linear model (CMLM) and CMLM-based genomic prediction and selection. The GAPIT package can handle large datasets in excess of 10 000 individuals and 1 million single-nucleotide polymorphisms with minimal computational time, while providing user-friendly access and concise tables and graphs to interpret results. AVAILABILITY http://www.maizegenetics.net/GAPIT. CONTACT zhiwu.zhang@cornell.edu SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Maize HapMap2 identifies extant variation from a genome in flux

Whereas breeders have exploited diversity in maize for yield improvements, there has been limited progress in using beneficial alleles in undomesticated varieties. Characterizing standing variation in this complex genome has been challenging, with only a small fraction of it described to date. Using a population genetics scoring model, we identified 55 million SNPs in 103 lines across pre-domestication and domesticated Zea mays varieties, including a representative from the sister genus Tripsacum. We find that structural variations are pervasive in the Z. mays genome and are enriched at loci associated with important traits. By investigating the drivers of genome size variation, we find that the larger Tripsacum genome can be explained by transposable element abundance rather than an allopolyploid origin. In contrast, intraspecies genome size variation seems to be controlled by chromosomal knob content. There is tremendous overlap in key gene content in maize and Tripsacum, suggesting that adaptations from Tripsacum (for example, perennialism and frost and drought tolerance) can likely be integrated into maize.

The Genetic Architecture of Maize Height

Height is one of the most heritable and easily measured traits in maize (Zea mays L.). Given a pedigree or estimates of the genomic identity-by-state among related plants, height is also accurately predictable. But, mapping alleles explaining natural variation in maize height remains a formidable challenge. To address this challenge, we measured the plant height, ear height, flowering time, and node counts of plants grown in >64,500 plots across 13 environments. These plots contained >7300 inbreds representing most publically available maize inbreds in the United States and families of the maize Nested Association Mapping (NAM) panel. Joint-linkage mapping of quantitative trait loci (QTL), fine mapping in near isogenic lines (NILs), genome-wide association studies (GWAS), and genomic best linear unbiased prediction (GBLUP) were performed. The heritability of maize height was estimated to be >90%. Mapping NAM family-nested QTL revealed the largest explained 2.1 ± 0.9% of height variation. The effects of two tropical alleles at this QTL were independently validated by fine mapping in NIL families. Several significant associations found by GWAS colocalized with established height loci, including brassinosteroid-deficient dwarf1, dwarf plant1, and semi-dwarf2. GBLUP explained >80% of height variation in the panels and outperformed bootstrap aggregation of family-nested QTL models in evaluations of prediction accuracy. These results revealed maize height was under strong genetic control and had a highly polygenic genetic architecture. They also showed that multiple models of genetic architecture differing in polygenicity and effect sizes can plausibly explain a population’s variation in maize height, but they may vary in predictive efficacy.

Genetic association mapping identifies single nucleotide polymorphisms in genes that affect abscisic acid levels in maize floral tissues during drought

In maize, water stress at flowering causes loss of kernel set and productivity. While changes in the levels of sugars and abscisic acid (ABA) are thought to play a role in this stress response, the mechanistic basis and genes involved are not known. A candidate gene approach was used with association mapping to identify loci involved in accumulation of carbohydrates and ABA metabolites during stress. A panel of single nucleotide polymorphisms (SNPs) in genes from these metabolic pathways and in genes for reproductive development and stress response was used to genotype 350 tropical and subtropical maize inbred lines that were well watered or water stressed at flowering. Pre-pollination ears, silks, and leaves were analysed for sugars, starch, proline, ABA, ABA-glucose ester, and phaseic acid. ABA and sugar levels in silks and ears were negatively correlated with their growth. Association mapping with 1229 SNPs in 540 candidate genes identified an SNP in the maize homologue of the Arabidopsis MADS-box gene, PISTILLATA, which was significantly associated with phaseic acid in ears of well-watered plants, and an SNP in pyruvate dehydrogenase kinase, a key regulator of carbon flux into respiration, that was associated with silk sugar concentration. An SNP in an aldehyde oxidase gene was significantly associated with ABA levels in silks of water-stressed plants. Given the short range over which decay of linkage disequilibrium occurs in maize, the results indicate that allelic variation in these genes affects ABA and carbohydrate metabolism in floral tissues during drought.

Iterative Usage of Fixed and Random Effect Models for Powerful and Efficient Genome-Wide Association Studies

False positives in a Genome-Wide Association Study (GWAS) can be effectively controlled by a fixed effect and random effect Mixed Linear Model (MLM) that incorporates population structure and kinship among individuals to adjust association tests on markers; however, the adjustment also compromises true positives. The modified MLM method, Multiple Loci Linear Mixed Model (MLMM), incorporates multiple markers simultaneously as covariates in a stepwise MLM to partially remove the confounding between testing markers and kinship. To completely eliminate the confounding, we divided MLMM into two parts: Fixed Effect Model (FEM) and a Random Effect Model (REM) and use them iteratively. FEM contains testing markers, one at a time, and multiple associated markers as covariates to control false positives. To avoid model over-fitting problem in FEM, the associated markers are estimated in REM by using them to define kinship. The P values of testing markers and the associated markers are unified at each iteration. We named the new method as Fixed and random model Circulating Probability Unification (FarmCPU). Both real and simulated data analyses demonstrated that FarmCPU improves statistical power compared to current methods. Additional benefits include an efficient computing time that is linear to both number of individuals and number of markers. Now, a dataset with half million individuals and half million markers can be analyzed within three days.

Simulation Appraisal of the Adequacy of Number of Background Markers for Relationship Estimation in Association Mapping

Complex trait dissection through association mapping provides a powerful complement to traditional linkage analysis. The genetic structure of an association mapping panel can be estimated by genomewide background markers and subsequently accounted for in association analysis. Deciding the number of background markers is a common issue that needs to be addressed in many association mapping studies. We first showed that the adequacy of markers in relationship estimation influences the maximum likelihood of the model explaining phenotypic variation and demonstrated this influence with a series of computer simulations with different trait architectures. Analyses and computer simulations were then conducted using two different data sets: one from a diverse set of maize (Zea mays L.) inbred lines with a complex population structure and familial relatedness, and the other from a group of crossbred dogs. Our results showed that the likelihood‐based model‐fitting approach can be used to quantify the robustness of genetic relationships derived from molecular marker data. We also found that kinship estimation was more sensitive to the number of markers used than population structure estimation in terms of model fitting, and a robust estimate of kinship for association mapping with diverse germplasm requires a certain amount of background markers (e.g., 300–600 biallelic markers for the simulated pedigree materials, >1000 single nucleotide polymorphisms or 100 simple sequence repeats [SSRs] for the diverse maize panel, and about 100 SSRs for the canine panel). Kinship construction with subsets of the whole marker panel and subsequent model testing with multiple phenotypic traits could provide ad hoc information on whether the number of markers is sufficient to quantify genetic relationships among individuals.

Quantitative trait loci for hip dysplasia in a crossbreed canine pedigree

Canine hip dysplasia is a common developmental inherited trait characterized by hip laxity, subluxation or incongruity of the femoral head and acetabulum in affected hips. The inheritance pattern is complex and the mutations contributing to trait expression are unknown. In the study reported here, 240 microsatellite markers distributed in 38 autosomes and the X chromosome were genotyped on 152 dogs from three generations of a crossbred pedigree based on trait-free Greyhound and dysplastic Labrador Retriever founders. Interval mapping was undertaken to map the QTL underlying the quantitative dysplastic traits of maximum passive hip laxity (the distraction index), the dorsolateral subluxation score, and the Norberg angle. Permutation testing was used to derive the chromosome-wide level of significance at p < 0.05 for each QTL. Chromosomes 4, 9, 10, 11 (p < 0.01), 16, 20, 22, 25, 29 (p < 0.01), 30, 35, and 37 harbor putative QTL for one or more traits. Successful detection of QTL was due to the crossbreed pedigree, multiple-trait measurements, control of environmental background, and marked advancement in canine mapping tools.

Fine Quantitative Trait Loci Mapping of Carbon and Nitrogen Metabolism Enzyme Activities and Seedling Biomass in the Maize IBM Mapping Population

Understanding the genetic basis of nitrogen and carbon metabolism will accelerate the development of plant varieties with high yield and improved nitrogen use efficiency. A robotized platform was used to measure the activities of 10 enzymes from carbon and nitrogen metabolism in the maize (Zea mays) intermated B73 × Mo17 mapping population, which provides almost a 4-fold increase in genetic map distance compared with conventional mapping populations. Seedling/juvenile biomass was included to identify its genetic factors and relationships with enzyme activities. All 10 enzymes showed heritable variation in activity. There were strong positive correlations between activities of different enzymes, indicating that they are coregulated. Negative correlations were detected between biomass and the activity of six enzymes. In total, 73 significant quantitative trait loci (QTL) were found that influence the activity of these 10 enzymes and eight QTL that influence biomass. While some QTL were shared by different enzymes or biomass, we critically evaluated the probability that this may be fortuitous. All enzyme activity QTL were in trans to the known genomic locations of structural genes, except for single cis-QTL for nitrate reductase, Glu dehydrogenase, and shikimate dehydrogenase; the low frequency and low additive magnitude compared with trans-QTL indicate that cis-regulation is relatively unimportant versus trans-regulation. Two-gene epistatic interactions were identified for eight enzymes and for biomass, with three epistatic QTL being shared by two other traits; however, epistasis explained on average only 2.8% of the genetic variance. Overall, this study identifies more QTL at a higher resolution than previous studies of genetic variation in metabolism.