Genome-Wide Association Mapping and Genomic Prediction Analyses Reveal the Genetic Architecture of Grain Yield and Agronomic Traits under Drought and Optimum Conditions in Maize

doi:10.21203/rs.3.rs-5289238/v1

Genome-Wide Association Mapping and Genomic Prediction Analyses Reveal the Genetic Architecture of Grain Yield and Agronomic Traits under Drought and Optimum Conditions in Maize

2024 · doi:10.21203/rs.3.rs-5289238/v1

preprint OA: closed

Full text JSON View at publisher

Full text 236,751 characters · extracted from preprint-html · click to expand

Genome-Wide Association Mapping and Genomic Prediction Analyses Reveal the Genetic Architecture of Grain Yield and Agronomic Traits under Drought and Optimum Conditions in Maize | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Genome-Wide Association Mapping and Genomic Prediction Analyses Reveal the Genetic Architecture of Grain Yield and Agronomic Traits under Drought and Optimum Conditions in Maize Manigben Kulai Amadu, Yoseph Beyene, Vijay Chaikam, Pangirayi B. Tongoona, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5289238/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 01 Feb, 2025 Read the published version in BMC Plant Biology → Version 1 posted 4 You are reading this latest preprint version Abstract Background Drought is a major abiotic stress in sub-Saharan Africa, impacting maize growth and development leading to severe yield loss. Drought tolerance is a complex trait regulated by multiple genes, making direct grain yield selection ineffective. To dissect the genetic architecture of grain yield and flowering traits under drought stress, a genome-wide association study (GWAS) was conducted on a panel of 236 maize lines testcrossed and evaluated under managed drought and optimal growing conditions in multiple environments using seven multi-locus GWAS models (mrMLM, FASTmrMLM, FASTmrEMMA, pLARmEB, pKWmEB, ISIS EM-BLASSO, and FARMCPU) from mrMLM and GAPIT R packages. Genomic prediction with RR-BLUP model was applied on BLUEs across locations under optimum and drought conditions. Results A total of 172 stable and reliable quantitative trait nucleotides (QTNs) were identified, of which 77 are associated with GY, AD, SD, ASI, PH, EH, EPO and EPP under drought and 95 are linked to GY, AD, SD, ASI, PH, EH, EPO and EPP under optimal conditions, with 17 QTNs explaining over 10% of the phenotypic variation ( R 2 ≥ 10%). Furthermore, 43 candidate genes were discovered and annotated. Two major candidate genes, Zm00001eb041070 closely associated with grain yield near peak QTN, qGY_DS1.1 (S1_216149215) and Zm00001eb364110 closely related to anthesis-silking interval near peak QTN, qASI_DS8.2 (S8_167256316) were identified, encoding AP2-EREBP transcription factor 60 and TCP-transcription factor 20, respectively under drought stress. Haplo-pheno analysis identified superior haplotypes for qGY_DS1.1 (S1_216149215) associated with the higher grain yield under drought stress. Genomic prediction revealed moderate to high prediction accuracies under optimum and drought conditions. Conclusion The lines carrying superior haplotypes can be used as potential donors in improving grain yield under drought stress. Integration of genomic selection with GWAS results leads not only to an increase in the prediction accuracy but also to validate the function of the identified candidate genes as well increase in the accumulation of favorable alleles with minor and major effects in elite breeding lines. This study provides valuable insight into the genetic architecture of grain yield and secondary traits under drought stress. Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Maize ( Zea mays L.) is an indispensable cereal crop in global agri-food systems[ 1 , 2 ]. However, grain yield is stagnating due to unpredictable climate change and increase in negative impacts of drought on maize production and productivity in sub-Saharan Africa (SSA) [ 3 – 6 ]. Boosting maize grain yield potential and improving maize resilience to drought are key solutions proposed for mitigating the effects of drought and climate changes while minimizing farmer’s risk [ 7 , 8 ]. In SSA, it is estimated that about 40% of the region's maize-growing area experiences intermittent drought stress, leading to sizable yield reductions ranging from 10–25% [ 9 , 10 ]. Maize plants require at least 500 to 800 mm of water, which translates to 5.6–6.7 mm per day. This largely depends on the maturity group, soil moisture, growth stage, and environmental conditions [ 11 , 12 ]. Water deficit below this range can lead to severe water stress in maize plants, particularly during the flowering stage. The maize plants respond to drought stress by rolling their leaves, reducing leaf area, and closing stomata, which affects photosynthetic activity and enzyme production. Drought stress coinciding with the flowering and grain-filling stages in maize causes a sizable yield reduction of up to 90% [ 13 – 15 ]. Considerable efforts have been made to boost the grain yield potential and stress-resilience in maize through conventional breeding[ 3 , 16 , 17 ]. However, genetic improvement of grain yield under drought stress through conventional breeding methods has proven to be challenging [ 18 ], primarily caused by the multigenic nature of traits, controlled by many loci, each contributing a small effect[ 13 , 19 , 20 ]. Owing to its multigenic nature of inheritance and genotype x environment interactions, grain yield often presents a low heritability under drought conditions [ 21 ]. For this reason, it is difficult to accurately estimate breeding values, which results in lower genetic gain per unit time and thereby constrains the development of drought-tolerant maize hybrids[ 22 ]. This, therefore, emphasizes the need to complement conventional breeding methods with genomic-assisted breeding tools to accelerate the development of high yielding drought-tolerant maize cultivars thereby boosting productivity in stress-prone areas. Recent advances in crop genomics and phenomics have increased our understanding of the physiological and genetic basis of complex traits such as drought tolerance and grain yield [ 20 , 23 ]. Linkage mapping based on biparental population is one of the most powerful tools extensively utilized to identify several quantitative trait loci (QTL) related to grain yield and secondary traits under drought stress [ 22 , 24 , 25 ]. However, many constructed maps suffer from low resolution and low allele richness [ 25 ]. Contrary to linkage mapping, GWAS utilizes diverse natural populations, which eliminates the need for developing segregating populations, saving time and cost[ 26 ] and hence is the most preferred method [ 27 , 28 ]. Moreover, it can detect multi-allelic variation, rare, and small effect QTLs simultaneously [ 29 ], providing high resolution by leveraging historical meiotic recombination events available in diverse natural populations[ 30 – 32 ]. Conversely, the results of GWAS can be influenced by the population structure, which can significantly interfere with the power of QTL detection[ 31 , 33 ]. For this reason, several statistical models have been developed including single-locus GWAS models and multi-locus GWAS models [ 33 ]. In single-locus GWAS models such as the general linear model (GLM) [ 34 ]; Mixed linear model (MLM)[ 35 ]; Enriched compressed mixed linear model (ECMLM)[ 36 ]; efficient mixed model association eXpedited (EMMAX) [ 37 ] and genome-wide efficient mixed-model association (GEMMA)[ 38 ], test marker-trait associations for significance by multiple testing one marker at a time. These models incorporate population structure (e.g. Principal component, Kinship matrices, etc.) as fixed covariates or a random polygenic effect to address the genetic relatedness among individuals in a diverse population [ 39 ] and tend to detect major QTLs. However, single locus models are often prone to high false positive rates or Type 1 errors. Bonferroni correction is commonly employed to control false positive rate (FPR)[ 40 ]. However, the use of Bonferroni correction has been proven to be too conservative such that true quantitative trait nucleotides (QTNs) may be missed out when considering SNPs in linkage disequilibrium (LD). Therefore, multi-locus GWAS models have been recommended for addressing multiple test corrections[ 41 ]. Multi-locus GWAS models do not require Bonferroni correction, have higher power for detection of both major and minor QTL effects, and have proven to be superior to single locus models in detecting small effect loci[ 40 ]. For this reason, several multi-locus GWAS including the multiple loci mixed model (MLMM), fixed and random model circulating probability unification (FarmCPU)[ 42 ], and Bayesian-information and linkage-disequilibrium iteratively nested keyway (BLINK)[ 43 ]. In addition, other key models are multi-locus random-SNP-effect mixed linear model (mrMLM)[ 44 ], the fast multi-locus random-SNP-effect mixed linear model (FASTmrMLM) [ 45 ], the fast multi-locus random-SNP-effect efficient mixed-model association (FASTmrEMMA) [ 46 ], polygenic-background-control-based least angle regression plus empirical Bayes (pLARmEB)[ 47 ], polygenic-background-control-based Kruskal–Wallis test plus empirical Bayes (pKWmEB) [ 48 ], and Iterative sure independence screening expectation maximization Bayesian least absolute shrinkage and selection operator (ISIS EM-BLASSO) [ 49 ] are commonly used now. The multi-locus GWAS methods involve a two-step process. In the first step, a single-dimensional genome scan is implemented using a less stringent critical value to identify putative QTLs. In the second step, all putative QTLs identified in the first step are subjected to further genome-wide scan to identify true QTNs using a logarithm of the odds (LOD) statistics to determine their significance[ 50 ]. Genomic selection (GS) or genomic prediction (GP) is another efficient genomic-assisted breeding tool in which genome-wide markers are fed into a prediction model to predict the genomic estimated breeding values (GEBVs) of lines in a breeding population [ 51 ]. It has enormous potential for improving drought tolerance in maize, as it allows for more accurate selection of complex traits such as grain yield under drought and reduces the breeding cycles by enhancing the genetic gain per unit time [ 52 ]. Several studies have demonstrated the effectiveness of utilizing GS for improving grain yield and secondary traits in drought tolerance breeding in maize[ 7 , 53 – 55 ]. Zhang et al.[ 56 ] reported low to medium prediction accuracy for grain yield (GY) and secondary traits in maize under drought stress. The results indicated that several factors influenced the prediction accuracy values, including the types of breeding populations, the size of the training population, the complex nature of traits, the marker densities, and the genotyping platforms. Therefore, the combined application of different GWAS models and GS can enhance the power of QTL detection and accelerate breeding to improve drought tolerance and assist in selecting superior genotypes under drought stress. The objectives of the study were to (1) identify significant QTNs and putative candidate genes for GY and secondary traits under optimum and drought stress (DS) using multi-locus GWAS models and (2) assess the potential of GS in improving GY and related traits under DS and optimum conditions. These results will further deepen our understanding of the genetic architecture of complex traits, especially for GY and drought tolerance, which is critical for making accurate selections and developing stress-resilient maize hybrids. Material and Methods Germplasm, Experimental design, and Phenotyping A panel of 236 maize inbred lines was assembled for this study. The lines were developed by International Maize and Wheat Improvement Centre’s (CIMMYT) Global Maize Program through conventional breeding and doubled haploid technology [ 57 ]. These elite maize lines were crossed with a popular single cross hybrid tester (CML566 x CML395) and produced 236 test cross hybrids. All test cross hybrids plus six commercial hybrids as checks (DK8031, H513, PH3253, Pioneer 3253, and WH505) were evaluated under DS and optimum conditions (Manigben et al., in review). The optimum experiments were conducted under rainfed conditions which was augmented with irrigation to avoid DS at seven (7) locations, that is, Embu [0.48° S 37.47° E, 1159 masl], Kaguru [37.67°E,-0.08°S,1463masl], Kakamega [0.29°N, 34.77°E,1535 masl], Kiboko[37.72°E, 2.22°S, 975 masl], Kirinyaga [37.19`E, 0.34`S, 1282masl], Mwtapa [3.93° S 39.74° E, 30 masl] and Shikutsa [ 0.28° N, 34.75° E ,1561 masl]. The drought experiment was conducted at three (3) locations, namely Kiboko [37.72°E, 2.22°S, 975 masl], Homabay [0.52°S, 34.45°E, 1751 masl] and Mtwapa [3.93°S, 39.74°E, 30 masl] during the dry season. The drought experiment was conducted following the protocol established by CIMMYT [ 19 , 58 ]. The drought experiments were irrigated once a week using a drip irrigation system until two weeks before the expected flowering date. Irrigation was withdrawn to maintain DS until harvest. The experiments were set up in a 5 x 49 alpha lattice design with two replications. The experimental unit was a two-row plot of 5m long with intra-row spacing of 0.25 m and inter-row spacing of 0.75 m. Two seeds per hill were seeded and later thinned to one plant per hill three weeks after seedling emergence to a final plant population of 53,333 plants/ha. Basal fertilizer application was carried out at planting using di-ammonium phosphate (D.A.P) fertilizer at the rate of 60 Kg N and 60 Kg P 2 O 5 per hectare. Six weeks after emergence, all experiments were top-dressed with Urea at the rate of 60 Kg N. All the experiments were kept weeds-free by manual weeding and herbicide control. Detailed information on the pedigree of the inbred lines used in this study is presented in Supplementary Table S1 . Phenotypic data collection and analysis In all the experiments, data were collected on days to 50% anthesis (AD) and days to 50% silking (SD) as the number of days from planting to the day, when half of the plants per plot had shed pollen and silks emerged, respectively. Anthesis-silking interval (ASI) was computed as the difference between SD and AD. Plant height (PH) and ear height (EH) were measured in centimeters from the base of the plant to the height of the first tassel branch and the node bearing the upper ear, respectively. Ear position (EPO) was measured as the ratio of ear height to plant height per plot. The number of ears per plant (EPP) was determined by dividing the total number of ears per plot by the number of plants harvested per plot. Ears from each plot were shelled and weighed to determine grain yield (GY) in kilograms, which was then converted to tons per hectare (t/ha). Moisture content (MOI) of the shelled grains at harvest was measured with a portable handheld moisture meter and recorded in percentage. GY per plot in tons per hectare will be calculated using the field weight of harvested ears per plot and adjusted 12.5% moisture content. All trait measurements were done according to the procedure outlined in the drought phenotyping protocol of CIMMYT [ 58 , 59 ]. Data was analyzed for each and across locations under optimum and DS conditions. The restricted maximum likelihood (REML) estimates of variance components, coefficient of variation, broad-sense heritability, phenotypic and genetic correlation among traits for individual and combined analysis (Eq. 1 ) were estimated using multi-environment trial analysis R package (META-R) [ 60 ]. The linear mixed models available in META-R were implemented using the Lme4 R-package [ 61 ]. In the model, all factors were treated as random effects in this analysis except the genotype effect to estimate the best linear unbiased estimates (BLUEs). The best linear unbiased predictions (BLUPs) and variance components were estimated by treating all factors as random except replication and environments. $$\:{\varvec{Y}}_{\varvec{i}\varvec{j}\varvec{k}\varvec{l}}=\varvec{\mu\:}+\:{\varvec{E}\varvec{n}\varvec{v}}_{\varvec{i}}+{\varvec{R}\varvec{e}\varvec{p}}_{\varvec{j}}\left({\varvec{E}\varvec{n}\varvec{v}}_{\varvec{i}}\right)+{\varvec{B}\varvec{l}\varvec{o}\varvec{c}\varvec{k}}_{\varvec{k}}\left({\varvec{R}\varvec{e}\varvec{p}}_{\varvec{j}}{\varvec{E}\varvec{n}\varvec{v}}_{\varvec{i}}\right)+{\varvec{G}\varvec{e}\varvec{n}}_{\varvec{l}}+\:{\varvec{E}\varvec{n}\varvec{v}}_{\varvec{i}}\:x\:{\varvec{G}\varvec{e}\varvec{n}}_{\varvec{l}}+\:+{\varvec{\epsilon\:}}_{\varvec{i}\varvec{j}\varvec{k}\varvec{l}}$$ 1 where $\:{\varvec{Y}}_{\varvec{i}\varvec{j}\varvec{k}\varvec{l}}$ is the trait of interest, $\:\varvec{\mu\:}$ is the overall mean effect; $\:{\varvec{E}\varvec{n}\varvec{v}}_{\varvec{i}}$ is the effect of $\:{i}^{th}$ environment; $\:{\varvec{R}\varvec{e}\varvec{p}}_{\varvec{j}}\left({\varvec{E}\varvec{n}\varvec{v}}_{\varvec{i}}\right)$ is the $\:{j}^{th}$ replication within $\:{i}^{th}$ environment; $\:{\varvec{B}\varvec{l}\varvec{o}\varvec{c}\varvec{k}}_{\varvec{k}}\left({\varvec{R}\varvec{e}\varvec{p}}_{\varvec{j}}{\varvec{E}\varvec{n}\varvec{v}}_{\varvec{i}}\right)$ is the effect of $\:{k}^{th}$ incomplete block within $\:\:{j}^{th}$ replication in $\:{i}^{th}$ environment; $\:{\varvec{G}\varvec{e}\varvec{n}}_{\varvec{l}}$ is effect of $\:{l}^{th}$ genotype; $\:{\varvec{E}\varvec{n}\varvec{v}}_{\varvec{i}}\:x\:{\varvec{G}\varvec{e}\varvec{n}}_{\varvec{l}}$ is the effect of genotype x environment interactions and $\:{\varvec{\epsilon\:}}_{\varvec{i}\varvec{j}\varvec{k}\varvec{l}}$ is the residual effect. The variance components from the combined analysis were used to compute broad sense heritability [ 62 ]. DNA extraction, Sequencing, SNP discovery, and calling. Genomic Deoxyribonucleic acid (DNA) of 236 lines was extracted from seedlings at the 4-leaf stage using a modified version of CIMMYT’s high throughput mini-prep Cetyl Trimethyl Ammonium Bromide (CTAB) protocol [ 63 ]. The DNA samples were shipped to Cornell University for genotyping. In brief, the high-quality DNA extracted from each leaf sample of the 236 lines was digested with restriction endonuclease Ape KI. DNA libraries were constructed for each sample and sequenced using genotyping-by-sequencing (GBS) protocol as described by Elshire et al. [ 64 ]. Raw GBS data of 955,690 SNPs distributed across the ten maize chromosomes were received from the Institute of Biotechnology at Cornell University, USA, after mapping to B73 AGPv2 coordinates. SNP calling was carried out using the TASSEL-GBS pipeline [ 65 ]. The raw GBS data were cleaned by removal of SNP markers with a minimum count of 90%, greater than 5% heterozygosity, and less than 5% minor allele frequency using TASSEL software version 5.2 [ 66 ], resulting in a total of 230,743 SNPs. In addition, the lines with greater than 20% missing data and SNPs not located on any of the ten chromosomes were further filtered out to a final dataset of 215,542 SNPs in 236 diverse lines for further analysis. The density and distribution map of SNPs on each of the ten maize chromosomes was drawn using a CMplot R package ( https://github.com/YinLiLin/CMplot ). Population structure and linkage disequilibrium analysis To capture population structure and cryptic genetic relatedness among the 236 lines, population structure, and kinship analysis were carried out using 215,542 genome-wide SNPs distributed across the ten chromosomes. The population structure was estimated using the admixture model method implemented in the software package STRUCTURE version 2.3.4 [ 67 , 68 ]. The number of subpopulations (K) was set from 1 to 15 with 10 independent runs for each K. The burn-in length and Markov Chain Monte Carlo (MCMC) replication were set at 100,000 each run under the admixture and correlated allele frequency model. The STRUCTURE HARVESTER [ 69 ], a web-based program was used to summarize STRUCTURE output, visualize the likelihood values across multiple values of K and, compute the natural logarithms of probability data [LnP(K) ] and the ad hoc statistic ΔK based on Evanno method [ 70 ]. The principal component analysis (PCA) was conducted using the Genomic Association and Prediction Integrated Tool (GAPIT) version 3[ 71 , 72 ] to detect the subpopulation structure present in the panel. The kinship matrix was estimated using the VanRaden algorithm [ 73 ] to measure the genetic relatedness among individuals in the association panel. The genetic relationship among the lines was determined based on the neighbor joining tree algorithm using the phylogenetic tree analysis in TASSEL software v5.2.93 [ 66 ]. To determine the extent of linkage disequilibrium (LD), squared allele frequency correlations ( r 2 ) between all pairs of SNP markers were estimated using TASSEL software version 5.2.93 [ 66 ]. To calculate the LD decay rate, the nonlinear regression model developed by [ 74 ], with modifications by Remington et al [ 75 ], was used to fit the LD decay curve into the scatterplot using the LOESS function in R. Genome-Wide Association Study GWAS analysis was carried out with 215,542 high quality SNPs (G) from 236 lines with BLUP values of eight phenotypic traits (GY, AD, SD, ASI, EPO, EPP, PH, and EH) using different multi-locus (ML) GWAS models under DS and optimum conditions. The first four PCAs and kinship (K) matrix were incorporated in the GWAS models as covariates to reduce false positives. The ML-GWAS was conducted with seven models including: (1) Fixed and random model circulating probability unification (FarmCPU)[ 42 ], (2) mrMLM[ 44 ], (3) FASTmrMLM [ 45 ], (4) FASTmrEMMA [ 46 ], (5) pLARmEB[ 47 ], (6) pKWmEB[ 48 ], and (7) ISIS EM-BLASSO[ 45 ]. One multi-locus model was implemented in (GAPIT) R package software[ 71 , 72 ] and the other six multi-locus models were implemented in the mrMLM R package[ 40 ]. The nomenclature for naming QTN was done using the letter "q" to indicate QTN, followed by an abbreviation representing the trait name, underscoring the management conditions, the corresponding chromosome number, and the number of QTNs identified on that specific chromosome. To determine the genome-wide significant P values threshold for the FarmCPU GWAS results, the effective number of independent SNPs (N) were calculated using the SimpM R program[ 30 , 76 , 77 ] available on GitHub ( https://github.com/LTibbs/SimpleM ). The genome-wide significant P value threshold was adjusted based on Bonferroni correction as the ratio of alpha value (α = 0.05) divided by the effective number of independent SNPs (N = 79,455). Hence, the genome-wide significant and suggestive levels were set as P = 0.05/N = 6.29 × 10 − 7 and P = 1/N = 1.26 × 10 − 5, respectively, where N is the effective number of independent SNPs. For multi-locus GWAS analysis, the genome-wide significant threshold was defined based on the threshold of LOD ≥ 3 ( p = 0.0002). Manhattan plots and quantile-quantile plots were developed to visualize GWAS results using CMplot R package ( https://github.com/YinLiLin/CMplot ). Candidate gene annotation and haplotype block analysis The LD decay with a physical distance of 4.75 kb found in this study was used to find candidate genes. All the candidate genes for GY, AD, SD, PH, EH, EPO, and EPP located within regions from 4.75 kb upstream to 4.75 kb downstream associated with significant QTNs were identified and annotated using the B73 maize reference genome (B73 RefGen_V2)[ 78 – 80 ]. The candidate gene annotations information was retrieved from the maizeGDB database ( http://www.maizegdb.org ). The significant QTN, qGY_DS1.1 (S1_216149215) on chromosome 1 for GY located within the genomic regions of a candidate gene, Zm00001eb041070 were extracted from the variant call format (VCF) using the site filtering option of VCF tools [ 81 ]. The haplotype block analysis was then implemented in Haploview software version 4.2 [ 82 ] and geneHapR [ 83 ]. The blocks were defined according to the criteria described by Gabriel et al. [ 84 ]. One-way Analysis of variance (ANOVA), boxplot, and multiple comparisons of phenotypic differences among haplotypes were implemented in agricolae R package using Tukey’s Honestly Significant differences (HSD) test. Genomic-wide prediction GP was carried out on the 236 lines based BLUE values for traits across environments within management using ridge regression best linear unbiased prediction ( rrBLUP ) model available in the rrBLUP R package [ 85 ]. Genomic estimated breeding values (GEBVs) were estimated using a five-fold cross-validation scheme by randomly sampling 80% and 20% of maize lines as training and testing sets, respectively. The prediction accuracy of the model was computed as the average Pearson’s correlation coefficient (r) between GEBV estimates from the training and testing set with 100 iterations. Bar plot was generated for each trait to visualize the means and standard deviation of prediction accuracy using the ggplot2 R package [ 86 ]. Results Phenotypic variation, Descriptive statistics, and Correlation The combined analysis of variance revealed significant ( p <0.05) genotype and GEI variations for GY, AD, SD, ASI, EPO, EPP, EH, and PH under DS and optimum conditions (Table 1). The mean, minimum, and maximum values of GY, AD, SD, ASI, EPO, EPP, EH, and PH revealed large variability for each trait under both DS and optimum conditions (Table 1). GY under DS was reduced by 71%. PH (214.69 cm) and EH (116.88 cm) were reduced significantly under DS compared to the mean of PH (225.94 cm) and EH (112.27 cm) under optimum conditions. DS had a relatively small effect on EPO and EPP. DS conditions increased the average number of days for AD and SD compared to optimum conditions. Consequently, prolonging the interval between AD and SD (known as ASI) under drought by an average of 3 days compared to optimum conditions. The broad-sense heritability ranged from 0.25 for GY to 0.87 for EPO under DS and varied from 0.21 for EPP to 0.74 for AD under optimum conditions. The frequency distribution of each of the traits under the two experimental conditions (DS and optimum) is shown in Figure 1 where most of the traits showed continuous distributions. Pearson’s correlation analysis showing the relationships among traits under DS and optimum conditions is presented in Figure 2. The correlation coefficients among the eight traits ranged from -0.29 to 0.93 under optimum conditions, whereas under DS conditions ranged from -0.56 to 0.86. Under optimum conditions, the highest significant positive coefficients correlation (r=0.93**) was observed between flowering traits, SD_OPT and AD_OPT, followed by PH_OPT and EH_OPT (r=0.78), GY_OPT and PH_OPT (r=0.65). GY_OPT had a significant positive correlation with PH_OPT, EH_OPT, and EPP_OPT while a significant but negative correlation was observed between GY_OPT with SD_OPT and ASI_OPT. Similarly, ASI_OPT had a significant negative correlation with AD_OPT, PH_OPT, EH_OPT, and EPP_OPT. Under DS conditions, the highest positive correlations were 0.83 (between GY_DS and EPP_DS), 0.66 (SD_DS and AD_DS), 0.64 (EPO_DS and EH_DS), 0.63 (PH_DS and EH_DS), and 0.57 (EPO_DS and AD_DS). GY_DS was negatively and significantly correlated with AD_DS, SD_DS, ASI_DS, and EPO_DS, while positively and significantly correlated with PH_DS and EPP_DS. It was observed that ASI_DS was positively correlated with AD_DS, SD_DS, and EPO_DS and negatively correlated with PH_DS, EH_DS, and EPP_DS. Additionally, AD_DS was positively correlated with SD_DS, EH_DS and EPO_DS, but negatively correlated with PH_DS and EPP_DS. Overall, a strong correlation was observed between the AD and SD under both optimum and DS conditions. Marker distribution, Population structure, Phylogenetic tree, and Kinship The distribution of 215,542 SNPs across the genome is presented in Figure 3A and supplementary Table S2. The number of SNP markers on each chromosome ranged from 14,629 to 33, 874 SNPs. Chromosome 1 had the highest number of SNPs of 33, 874, whereas chromosome 10 had the lowest number of SNPs with 14,629. The density of SNPs per mega-base pair (Mbp) varied from 84.89 Mbp for chromosome 4 to 116.79 Mbp for chromosome 5 with a mean of 104.28 Mbp. The population structure of 236 diverse maize lines was determined by Bayesian based model in STRUCTURE and PCA (Supplementary Figure S1a and Figure 3B). The optimum number of K was obtained by plotting the number of clusters (K) against delta K (Supplementary Figure S1a). The Bayesian structure analysis revealed the presence of three distinct subgroups within the 236 maize lines when K = 3 and ten distinct subgroups K = 10 (Supplementary Figure S1b). The genetic structure was further examined by PCA (Figure 3B). The results revealed the presence of four distinct subpopulations separated by PC1 (12.78 %), PC2 (9.11%), and PC3 (7.19 %). The first three PCs explained over 29 % of the total genetic variation among the diverse maize lines. A scree plot of variance explained against the corresponding PCs as shown in Figure 3C was used to determine the optimal number of PCs to retain. The scree plot revealed that an optimal number (K) of four PCs could be retained for GWAS. The phylogenetic tree based on the neighbor-joining method as shown in Figure 3d revealed that the 236 diverse maize lines can be clustered into four main groups (I=66, II=48, III=100, and IV=22) differentiated by the different colors (Supplementary Table S1). The kinship matrix is utilized to assess the relatedness among individuals by considering the extent of allele sharing. The pattern of red shading in the center of the kinship matrix (Figure 3F reflected the level of genetic relatedness among individuals, indicating the presence of four stratified population structures. In general, the results from the PCA, phylogenetic tree, and kinship matrix, show that the panel of 236 diverse maize lines can be divided into four subpopulations. Linkage Disequilibrium and GWAS analysis The nonrandom association of alleles between two genetic loci was examined to provide valuable information in locating genes tightly linked to SNP markers associated with traits of interest. A rapid LD decay pattern was observed across the ten chromosomes (Figure 3F). At r 2 = 0.2, the LD decay distance between SNPs on the 10 chromosomes varied from 2.83 kb (Chr3) to 11.83 kb (Chr4) with an average genome-wide LD decay with physical distance between SNPs of 4.74 kb. Based on the seven multi-locus GWAS models (FarmCPU, mrMLM, FastmrMLM, FASTmrEMMA, pLARmEB, pKWmEB and ISIS EM-BLASSO), 219 and 265 QTNs distributed across the 10 chromosomes were detected under DS and optimum conditions, respectively (Figure 4, Supplementary Table S3 and S4). The QQ-plots depicting the observed vs expected LOD [−log10 ( p ), p = 6.29e-07] distributions showed a significant deviation from the expected distribution (Supplementary Figure S2 and S3), indicating the presence of QTNs linked to GY, AD, SD, ASI, PH, EH, EPP and EPO, respectively under DS and optimum conditions. The Manhattan plot showed that several QTNs with Bonferroni-adjusted threshold of greater than 6.20 [−log10 ( p ), p = 6.29e-07] and LOD threshold of ≥ 3 (p = 0.0002), were associated with GY, AD, SD, ASI, PH, EH, EPP and EPO, respectively under DS and optimum conditions. The number of significant QTNs detected by multiple GWAS models varied across traits. Of these, the power of detection among the seven models followed FASTmrMLM (62 QTNs) > pLARmEB (45 QTNs) > ISIS EM-BLASSO (37 QTNs) > pKWmEB (36 QTNs) > FarmCPU (16 QTNs) > FASTmrEMMA (12 QTNs) > mrMLM (7 QTNs) under DS and FASTmrMLM (68 QTNs) > pLARmEB (62 QTNs) > pKWmEB (42 QTNs) > FASTmrEMMA (29 QTNs) > FarmCPU (25 QTNs) >ISIS EM-BLASSO (24 QTNs) > mrMLM (15 QTNs) under optimum conditions (Fig. 4C and 4D). Of these QTNs, 46, 33, 31, 30, 29, 20, 17, and 13 were found to be associated with EPO, ASI, AD, GY, SD, EPP, EH and PH under DS, respectively. Under optimum conditions, 40, 37, 36, 36, 32, 25, and 24 QTNs were associated with EPO, EH, ASI, GY, EPP, PH, SD and AD, respectively (Fig. 4A and 4B). With the proportion of R 2 (%) by individual QTNs accounting for 0.00 ~ 8.18, 0.60 ~ 16.06, 3.93 ~ 15.29, 1.43 ~30.68, 0.00 ~ 11.42, 0.75 ~ 10.08, 1.32 ~ 16.06 and 2.00 ~ 12.66 (%) of total phenotypic variation for EPO, EH, ASI, GY, EPP, PH, SD and AD, respectively (Supplementary Table S4, S5 and S6). Significant QTNs detected by at least two ML-GWAS models were considered reliable and stable. A sum of 172 QTNs comprising 77 QTNs under DS and 95 QTNs under optimum conditions were discovered to be significantly associated with EPO, ASI, AD, GY, SD, EPP, EH, and PH, respectively (Fig. 4E and 4F; Supplementary Table S6). Of these significant QTNs under DS, 9, 9, 9, 11, 6, 6, 7, and 20 reliable QTNs were detected for GY, AD, SD, ASI, PH, EH, EPP, and EPO, respectively. The highest R 2 (%) value was observed for GY QTN qGY_DS2.1 [S2_194907656: R 2 =5.07 %-11.07%], for AD qAD_DS1.2 [S1_32228028: R 2 =5.27%-11.27%], for SD qSD_DS10.1 [S10_2032146: R 2 = 5.36 % - 9.92 %], for ASI qASI_DS5.1 [S5_151886720: R 2 =5.94 %-11.19 %], for PH qPH_DS3.2 [S3_149683959: R 2 = 9.62%-11.27 %], for EH qEH_DS2.1 [S2_60655022: R 2 = 6.23%-6.78%], for EPP qEPP_DS1.1 [S1_217115118: R 2 = 9.19 %-13.26 %] and for EPO qEPH_DS2.1 [S2_36458273: R 2 = 0.96 % - 6.15 %]. In addition, of the 95 QTNs detected in optimum condition, 12, 9, 6, 13, 13, 11, 18, and 13 QTNs were found to be associated with GY, AD, SD, ASI, EH, PH, EPO, and EPP, respectively. The QTN, qGY_OPT5.2 [S5_193538664: R 2 = 5.34% - 10.19%] recorded the highest R 2 (%) for GY. One QTN, qAD_OPT8.1 (S8_105322358) was shared by AD ( R 2 = 6.03 % -12.66 %) and SD ( R 2 = 3.99 %-16.06 %). For ASI, qASI_OPT6.1 [S6_159006932: R 2 = 6.30% - 14.10%] explained the highest R 2 (%). While qEH_OPT10.4 [S10_88760138: R 2 = 9.39 % - 16.06 %] showed the highest R 2 (%) for EH, qPH_OPT3.1 [S3_172201680: R 2 = 1.49 % - 10.08%] for PH, qEPO_OPT2.2 [S2_233748945: R 2 = 1.56 % - 3.26 %] for EPO and qEPP_OPT8.2 [S8_75600223: R 2 = 0.002 % -5.48 %] for EPP (Figure 4 and Figure 5). The detected QTNs contributing revealed positive and negative allelic effects for different traits (Figure 4E and Figure 5). Among the 172 QTNs, few QTNs were detected under both DS and optimum conditions. For instance, two QTNs qGY_DS2.1 and qGY_DS1.1 explained the highest and positive effects with 0.34 and 0.24, respectively under DS for GY. On the other hand, two new QTNs qGY_OPT4.1 and qGY_OPT10.1 showed the highest and positive effects of 0.56 and 0.49, respectively for GY under optimum conditions. Negative QTN effects are desirable for flowering traits (AD, SD, and ASI) under DS for selecting earliness. For instance, QTNs qAD_DS2.1 , qSD_DS8.1 , qASI_DS1.1 exhibited large and negative effects for AD (-1.72), SD (-1.79), and ASI (-0.77) on chromosome 2, 8 and 1, respectively (Figure 4E). The distribution of QTNs (Figure 5) on the 10 chromosomes showed that chromosome 5 captured the higher number of reliable QTNs (12 under DS and 14 under optimum conditions). Two QTNs, qPH_OPT2.2 and qEH_OPT2.1 exhibited a pleiotropic effect on both PH and EH under optimum conditions. Candidate gene identification and Annotation Candidate genes analysis was conducted for significant QTNs identified in this study to elucidate the molecular, biological, and physiological mechanisms controlling traits under DS and optimum conditions. A total of 43 candidate genes were discovered and annotated, among them 18 and 25 candidate genes were identified under DS and optimum conditions, respectively (Table 2). Two candidate genes closely associated with the QTNs for improved GY were identified under DS, namely, Zm00001eb041070 and Zm00001d045665 (Table 2). Nine candidate genes potentially associated with flowering traits (AD, SD and ASI) were identified under DS (Table 2). Of these five candidate genes were associated with ASI under DS. Similarly, four and three candidate genes were found to be associated with EPO and EPP, respectively. Under optimum conditions, 3, 5, 8, 3, and 3 candidate genes were identified for GY, PH, EH, EPO, and EPP, respectively (Table 2). On the other hand, one candidate gene each was identified for AD, SD, and ASI. Haplotype Analysis The QTN qGY_DS1.1 (S1_216149215) associated with GY under DS was identified in a 216.15 Mb region on chromosome 1 based on pairwise LD correlation (Figure 6A and 6C). The QTN region contains six distinct SNPs with relatively high LD (Figure 6C). Four major haplotypes were detected among the 236 lines with Haplotypes frequencies of 142 (60 %), 39 (17 %), 29 (12 %), and 7 (3%) for Hap1 (GAGGGC), Hap2 (AAGGGC), Hap3 (GTAATG), and Hap4 (GAGGTG), respectively (Figure 6B). A significant difference was observed between haplotypes Hap1 and Hap2 (Figure 6D). The mean GY was 2.2 t/ha for Hap1, 1.80 t/ha for Hap2, 2.10 t/ha for Hap3, and 2.2 t/ha for Hap4, respectively. (Figure 6D). Hap1 was considered a superior haplotype since it contributes to the highest mean performance compared to other haplotypes. Genomic prediction Among the GP models, RR-BLUP is computationally less intensive and is well suited for routine application in plant breeding trials. Therefore, we used the RR-BLUP model to estimate the performance of maize genotypes for various traits under optimum and DS conditions (Figure 7). Prediction accuracies were moderate to high for all eight traits under optimum and low to moderate under DS conditions (Figure 7). The observed prediction accuracy for GY, AD, SD, ASI, PH, EH, EPO, and EPP were 0.29, 0.58, 0.45, 0.44, 0.61, 0.54, 0.24, and 0.25, respectively under DS conditions, and 0.65, 0.72, 0.69, 0.37, 0.71, 0.51, 0.48, and 0.41, under optimum conditions. Discussion Drought has long been recognized as a major abiotic factor limiting crop growth and productivity [ 4 , 10 , 87 ]. Genetic dissection of the genomic regions responsible for GY and other secondary traits under DS will allow breeders to improve their breeding efficiency in the development of climate-resilient varieties and, also, facilitate the introgression of the favorable alleles into elite germplasm using marker-assisted selection [ 88 ]. Furthermore, understanding the complex genetic basis of drought tolerance, GY and other secondary traits facilitate the opportunity to test for the indirect selection of GY under DS [ 25 , 89 ]. In the present study, significant genotypic variance and wide range phenotypic variation observed for all traits under optimal and DS, indicated the presence of adequate genetic diversity within the GWAS panel and that progress from selection for GY and secondary traits could be achieved through breeding. These findings support earlier studies [ 7 , 90 , 91 ], which indicated the existence of sufficient genetic variability. The notable significant GEI in the study implied that the testing sites, drought and optimal growing conditions were discriminating enough in identifying genotypic differences in the response of the GWAS panel to drought and optimal conditions, and that these differences can be largely attributed to the differences in environmental factors, such as soil types, temperature, and amount of rainfall [ 13 ]. Heritability estimates are crucial for ascertaining the effectiveness and progress anticipated from future phenotypic selection for yield and secondary traits under DS. The lower heritability estimates for GY under drought stress indicated that the correlated secondary trait, ASI with high heritability and strong correlation with GY could enhance the effectiveness of the selection response [ 92 ]. These findings are consistent with earlier studies [ 90 , 91 , 93 , 94 ]. A study by Ndlovu et al. [ 95 ] on multiple bi-parental maize populations in Kenya under water-stressed and well-watered conditions also reported lower heritability for GY and low genotypic variance under DS conditions. The notably significant and negative correlation between GY and ASI under DS indicated that ASI is a suitable secondary trait to facilitate the selection of GY and a target for improving drought tolerance in maize. These findings are consistent with earlier studies [ 94 , 96 , 97 ]. Linkage disequilibrium, population structure and association mapping The LD decay distance determines the power of the GWAS [ 98 ]. The high-resolution provided by GWAS is largely influenced by the nature of LD and the extent of its decay across the genome [ 29 , 30 ], which is population specific and influenced by recombination rate, number of generations of recombination, genetic drift, selection within populations, and population admixture [ 75 , 99 ]. In this study, genome-wide LD analysis revealed that the GWAS panel decayed rapidly at 4.75 kb (r 2 = 0.2). This suggests that the GWAS panel contained adequate genetic diversity suitable for GWAS analysis resulting from the historical and evolutionary recombination events present in the panel. Previous studies conducted by Rashid et al .[ 100 ], reported a rapid LD decay of 0.9 kb at r 2 = 0.2 for CIMMYT Asia Association Mapping, 1.75 kb at r 2 = 0.2 for Drought Tolerant Maize for Africa (DTMA) and 0.99 kb at r 2 = 0.2 for Improved Maize for African Soils (IMAS) panels. Lu et al. [ 101 ], reported that LD decay distance in temperate germplasm (10 to 100 kb), was two to ten times greater than LD decay distance in tropical germplasm (5 to 10 kb), suggesting that tropical maize has significant genetic diversity for breeding programs. The population structure within the GWAS panel was best explained when separated into four groups. The incorporation of the four PCs in GWAS analysis was sufficient to reduce false positive marker-trait association in the present study as evident in the Q-Q plots and Manhattan plots. The variation in the number of QTNs detected across the GWAS models suggests that the differences can be attributed to genetic algorithms implemented in the different models. A similar conclusion that multi-locus models show superior performance over single-locus models in terms of statistical power in detecting QTNs, particularly in the accuracy of QTN effect estimation and reducing false positive rates has been found in many other studies [ 33 , 41 , 102 , 103 ]. This further implied that adopting multiple GWAS models could help to complement each other in identifying reliable and stable QTNs [ 7 ]. The proportion of phenotypic variance explained R 2 (%) by each QTNs and the QTN effects observed for GY, EPP and EPO in this study were relatively low under DS. The results implied that the GY, EPP and EPO under DS are likely influenced by numerous QTNs with small effect. Similar results were reported in earlier study [ 104 ]. Candidate genes discovery In the present study, two candidate genes ( Zm00001eb041070 and Zm00001d045665 ) significantly associated with GY were identified under DS. The peak QTNs, qGY_DS1.1 is located within the gene, Zm00001eb041070 encoding AP2 (APETALA2)-EREBP (Ethylene Responsive Element Binding Protein) transcription factor 60 on chromosome 1. It plays a critical role in regulating genes involved in plant growth and development, biotic, and abiotic stress [ 105 ]. Ningning et al [ 106 ] reported three candidate genes, GRMZM2G322672 (EREB37), GRMZM2G026926 (ERF), and GRMZM2G169654 (RAV) , belonging to the AP2/EREBP family that plays important roles in maize response to DS. In maize, ZmEREBP60 is a positive regulator under DS and its expression, significantly upregulated in the roots, coleoptiles, and leaves in response to drought [ 107 , 108 ] Another gene Zm00001eb041070 reported to control AD under DS [ 104 , 109 ]. In addition, a gene Zm00001d045665 encoding Ubiquitin receptor Radiation sensitivity proteins-23 (RAD23) was identified − 46 bp around the peak QTN, qGY_DS.9.1 . This receptor is known to participate in DNA repair and play a major role in the cell cycle, morphology, and fertility of plants through their delivery of UPS substrates to the 26S proteasome [ 110 , 111 ]. The QTNs associated with GY, qGY_OPT2.1 , qGY_OPT1.1 , qGY_OPT10.1 are nearest to three candidate genes, Zm00001d005785, Zm00001eb034820 and Zm00001eb425230 which encode pentatricopeptide (PPR) repeat-containing protein [ 112 , 113 ], PfkB-like carbohydrate kinase family protein[ 114 ] and Trihelix-transcription factor (TTF) 25 [ 115 , 116 ], respectively. In a QTN-by environment interaction GWAS under optimum, DS, heat stress, and combined drought and heat stress, Wen et al [ 104 ] identified GRMZM2G110851 related to GY around the locus S1_299093763 encoding pentatricopeptide repeat protein, required for mitochondrial function and kernel development in maize. In rice and Arabidopsis, phosphofructokinase B-type carbohydrate kinase family protein, PFKB1 regulates leaf and plant growth by controlling chloroplast biogenesis[ 114 , 117 ]. The TTFs are light-responsive proteins and play a critical role in plant growth and development and stress tolerance in maize [ 116 , 118 ]. Flowering traits in maize are quantitative in nature and regulated by complex genetic mechanisms [ 119 ]. Several candidate genes associated with flowering traits under DS and optimal conditions have been reported in maize[ 24 , 54 , 104 , 120 , 121 ]. Here, some new candidates have been found to be associated with flowering traits. Rubisco large subunit methyltransferase encoded by Zm00001eb058200 on chromosome 1 near peak QTN qAD_DS1.1 was found under DS. It is a key photosynthetic enzyme in the chloroplast that catalyzes the fixation of atmospheric CO 2 during photosynthesis in the carbon assimilation pathway and drought-related responses [ 122 , 123 ]. The gene Zm00001eb010210 , associated with AD under DS encodes plant U-box type E3 ubiquitin ligase [ 124 , 125 ]. The ligase gene family plays a key role in the abiotic stress adaptation in maize [ 126 ]. Another candidate gene for AD, Zm00001eb012590 encodes Trifunctional UDP-glucose 46-dehydratase, an enzyme plays a key role in nucleotide sugar biosynthetic pathway, maintaining cell wall integrity during maize cell growth and response to abiotic stress[ 127 , 128 ]. ASI is a key index in breeding programs for facilitating indirect selection for GY under DS. Maize exhibits postponement in the silking process under stress, resulting in an extended period between AD and SD [ 89 ]. Five candidate genes Zm00001eb047160, Zm00001eb364110, Zm00001eb430770, Zm00001eb426690 and Zm00001d023240 associated with ASI under DS, encodes for kaurene synthase1[ 129 ], TEOSINTE BRANCHED1/CYCLOIDEA/PROLIFERATING CELL FACTOR (TCP) -transcription factor 20 [ 130 ], Ribosomal protein s11-beta [ 131 ], Hexosyltransferase[ 132 ] and Serine/threonine-protein kinase BLUS1[ 133 ], respectively. Additionally, the candidate gene, Zm00001eb047160 associated with PH under optimal conditions, encodes kaurene synthase, an enzyme responsible for the biosynthesis of gibberellins and phytoalexin metabolism [ 129 , 134 ]. The TCP family of transcription factors participates in the regulation of cell growth, proliferation, and response to abiotic stress [ 135 ]. In maize, ZmGA20ox3 play a crucial in regulating PH and enhancing drought tolerance [ 136 ]. Furthermore, candidate genes, Zm00001eb364110 and Zm00001eb408080 related to ASI and EPP on chromosome 8 and 10, respectively, under drought and optimal conditions encode TCP-transcription factor [ 137 , 138 ]. Previous studies reported that ZmTCP family proteins, particularly ZmTCP42 , is a key positive regulator of drought tolerance in maize. However, overexpression of the ZmTCP14 gene significantly reduced the tolerance to drought [ 138 ]. However, ZmTCP14 gene-edited plants showed improved drought tolerance and GY[ 137 , 138 ]. Additionally, the candidate gene Zm00001d023240 encodes serine/threonine-protein kinase, BLUE LIGHT SIGNALLING 1 (BLUS1) , play a crucial role in blue light-induced stomatal opening and transduction of signals related to abiotic stress [ 137 , 139 ]. This suggests that serine/threonine-protein kinase BLUS1 participates in regulating ASI during DS. Among the candidate genes identified with association to EPO under DS, Zm00001d044508 encodes cysteine-rich receptor kinase, which plays a critical role in plant immunity, and abiotic stress response [ 113 , 114 ]. Another gene Zm00001eb426230 encodes GDSL-motif esterase/acyltransferase/lipase, play a key role in cuticle development, seed oil storage, and biotic and abiotic stress responses [ 142 – 144 ]. The Zm00001d026371 gene associated with EPO encodes anthocyanidin reductase, a key enzyme in involved in mediating proanthocyanidin and lignin biosynthesis in response to DS [ 145 , 146 ]. Previous study [ 147 ] reported that the anthocyanidin reductase enzyme system plays a key role in reducing excess reactive oxygen species (ROS) induced by DS, which can damage cellular membranes and cause cell death in maize. The Zm00001d002247 gene, which underpinned an association with EPP under DS, encodes the MATH-BTB domain containing proteins, which play a key role in abiotic stress response [ 148 , 149 ]. The candidate gene Zm00001d001877 near peak QTN, qEH_OPT2.1 and qPH_OPT2.2 exhibited pleiotropic effect on EH and PH under optimal conditions. This implied that a positive correlation between these two traits and that the two QTNs simultaneously regulated [ 150 ]. The candidate gene, Zm00001eb431080 on chromosome 10 exhibited pleiotropic effect with PH and EPO under optimum conditions. This candidate gene encodes basic Helix-Loop-Helix ( bHLH ) transcription factor, involved in anthocyanin biosynthesis and response to biotic and abiotic stresses [ 151 – 153 ]. ABA-responsive protein encoded by Zm00001d023664 is associated with EPP plays a critical role in controlling stomatal closure, and regulate plants response to abiotic stress [ 154 , 155 ]. The candidate gene, Zm00001eb003210 associated with EPP encodes CONSTANS-LIKE (COL) TIMING OF CAB1 protein domain, which plays a crucial role in regulating flowering through photoperiodic control and response to abiotic stress [ 156 , 157 ]. Haplotype estimation Haplotype refers to a set of alleles combinations within a specific genomic region showing significant LD and are inherited together [ 83 , 158 ]. Identification of superior haplotypes linked to functional genes is important for developing markers to accelerate genetic gain and facilitate efficient selection in breeding operations. The haplo-pheno analysis identified Hap1 (GAGGGC) as a superior haplotype with high mean performance associated GY near QTN, qGY_DS1.1 ( S1_216149215 ) in the present study. Incorporating this superior haplotype in maize breeding will facilitate the selection of improved drought tolerant maize cultivars. These findings revealed that the GY of lines carrying superior haplotypes can exhibit improved drought tolerance compared to lines with unfavorable haplotypes. This is evident in their ability to maintain higher yields in DS and, these lines can also be utilized as parents for hybridization and backcrossing programs in the future. Genome-wide prediction GS has been successfully applied in maize for complex traits like GY under optimum, low soil N stress, heat stress and DS conditions [ 7 , 55 , 57 , 159 ]. In the present study, we compared the prediction accuracies under optimum and DS conditions (Fig. 7). The prediction accuracies were moderate to high for GY and other traits under optimum conditions whereas the accuracies were slightly lower under DS conditions. The accuracy observed for all traits under both optimum and DS conditions reveal the effect of heritability as the traits with higher heritability generally had higher prediction accuracy. The observed prediction accuracies for GY and other traits are comparable to earlier studies reported under different stresses in maize [ 5 , 7 , 57 , 95 , 159 , 160 ]. In genome-wide predictions, the less complex traits like AD, SD and PH had higher accuracy compared to the GY, which is consistent with the nature of trait complexity [ 7 , 55 , 56 ]. Breeding for drought tolerance is complex and very expensive. The observed prediction accuracy for GY is 0.29 and 0.65 under DS and optimum conditions, respectively. On the other hand, the phenotypic selection efficiency is 0.50 (square root of heritability of GY) under DS and 0.73 under optimum conditions. Nevertheless, with the possibility to complete three cycles per year and requirement of lesser resources for implementing GS compared to phenotypic selection, endorse GS to integrate to improve the selection efficiency for drought tolerance as an attractive option in a longer run objective of breeding program. Further, integration of GS with GWAS results leads not only to an increase in the prediction accuracy, but also helps to validate the function of the identified candidate genes and increases in the accumulation of favorable alleles for drought tolerance in improved set of elite lines. Further GS can remarkably reduce the resources required for selection and improve breeding efficiency. Conclusion Tremendous variations for GY and secondary traits associated with DS existed in the maize GWAS panel used in this study. Based on seven GWAS models, a total of 172 stable and reliable QTNs (77 under DS and 95 under optimal conditions) were identified. Among these QTNs two QTNs, qGY_DS1.1 for GY and qASI_DS8.2 for ASI were novel. This result demonstrated that combining different GWAS models is effective and powerful in exploiting the complementary strengths of different methods in identifying QTNs with small and large effects for GY and secondary traits with a complex genetic basis and low heritability under DS. In addition, a total of 43 candidate genes were detected (18 under DS and 25 optimum conditions). Among these, two key candidate genes are closely associated with GY and nine closely associated with flowering traits under DS. Genomic prediction revealed moderate to high accuracies under optimum and DS conditions. Integration of GS with GWAS results leads not only to an increase in prediction accuracy, but also helps to increase in the accumulation of favorable alleles for drought tolerance. Overall stable QTNs and candidate genes for GY and secondary trait detected in the current study can serve as invaluable resources for improving GY under DS in maize. The identified superior haplotype can be converted to a breeder-friendly marker to increase its favorable alleles in elite breeding lines. These findings provide valuable insight into the genetic basis of drought tolerance for GY and secondary traits in tropical maize. Abbreviations ANOVA: Analysis of variance AD: Anthesis days ASI: Anthesis-silking interval BLINK: Bayesian-information and linkage-disequilibrium iteratively nested keyway CV: Coefficient of variation CIMMYT: International Maize and Wheat Improvement Centre DS: Drought Stress ECMLM: Enriched compressed mixed linear model EMMAX: Efficient mixed model association eXpedited EPO: Ear position EH: Ear height EPP: Ear per plant FarmCPU: Fixed and random model Circulating Probability Unification FaST-LMM: Factored spectrally transformed linear mixed models FASTmrEMMA: Fast multi-locus random-SNP-effect efficient mixed model analysis FASTmrMLM: Fast mrMLM GP : Genomic prediction GS: Genomic selection GY: Grain yield GLM: General linear model GWAS: Genome-wide association study GEMMA: Genome-wide efficient mixed-model association ISIS EM-BLASSO: Iterative modified-sure independence screening expectation-maximization-Bayesian LASSO LD: Linkage disequilibrium LOD: Logarithm of Odds ML-GWAS: Multi-locus genome-wide association studies MLM: Mixed linear model mrMLM: multi-locus random-SNP-effect MLM PCA: Principal component analysis pLARmEB: polygenic-background-control-based least angle regression plus empirical Bayes pKWmEBb: Integration of Kruskal-Wallis test with empirical Bayes QTL: Quantitative trait locus QTN: Quantitative trait nucleotide R 2 : The proportion of total phenotypic variance explained by each QTN r 2 : The squared correlation coefficient SL-GWAS: Single-locus genome-wide association studies SNP: Single-nucleotide polymorphic Declarations Data Availability All data generated or analyzed during this study are included in this published article and its supplementary information files. All relevant data are within the manuscript and its supplementary files. Acknowledgments Our sincere thanks to CIMMYT’s research technicians in Kenya. The authors are grateful to the International Maize and Wheat Improvement Center (CIMMYT) and partner scientists and technicians who generated the germplasm and designed and conducted the experiments that we used to explore our objectives. Funding The research was supported by the Bill and Melinda Gates Foundation (B&MGF), Foundation for Food and Agriculture Research (FFAR), and the United States Agency for International Development (USAID) through AG2MW (Accelerating Genetic Gains in Maize and Wheat for Improved Livelihoods, B&MGF Investment ID INV-003439), the Stress Tolerant Maize for Africa (STMA, B&MGF Grant # OPP1134248) project, and the CGIAR Research Program on Maize (MAIZE). This research was part of the PhD thesis of the first author, funded by the WACCI ACE Impact project at West Africa Centre for Crop Improvement (WACCI), University of Ghana, Legon, Ghana. Authors and Affiliations International Maize and Wheat Improvement Center (CIMMYT), C/O: World Agroforestry Centre (ICRAF), United Nations Avenue, Gigiri, P.O. Box 1041-00621, Nairobi, Kenya. Manigben Kulai Amadu, Yoseph Beyene, Vijay Chaikam, Boddupalli M Prasanna and Manje Gowda 2 West Africa Centre for Crop Improvement (WACCI), University of Ghana, PMB 30 Legon, Accra, Ghana Manigben Kulai Amadu, Pangirayi B. Tongoona, Eric Y. Danquah and Beatrice E. Ifie CSIR-Savanna Agricultural Research Institute, PO. Box 52, Tamale, Nyankpala, Ghana Manigben Kulai Amadu, International Maize and Wheat Improvement Center (CIMMYT), Km 45, Carretera México-Veracruz, CP 52640, Edo. de México, Mexico Juan Burgueno Institute of Biological, Environmental & Rural Sciences (IBERS), Aberystwyth University, Aberystwyth, SY23 3EE, Wales, UK Beatrice E. Ifie Contributions Conceptualization, M.K.A., M.G., Y.B. and P.B.T.; methodology, M.G.,M.K.A., Y.B., V.C., and P.B.T; software, M.G., J.B, and M.K.A.; validation, M.K.A., M.G., Y.B., V.C., and P.B.T.; formal analysis, M.K.A.,M.G., J.B; investigation, Y.B., M.G. and M.K.A.; resources, B.M.P., M.G., E.Y.B. and V.C.; data curation, M.K.A., M.G. and Y.B.; writing—original draft preparation, M.K.A. and M.G.; writing—review and editing, M.K.A., M.G.,V.C, Y.B., B.M.P., J.B, P.B.T., and B.E.I., visualization, M.G. and Y.B.; supervision, M.G., P.B.T., E.Y.D., B.E.I. and B.M.P.; project administration, M.G. and Y.B.; funding acquisition, B.M.P and E.Y.D. All authors have read and agreed to the published version of the manuscript. Declarations Ethics approval and consent to participate All experimental studies on plants complied with relevant institutional, national, and international guidelines and legislation. Consent for publication Not applicable. References Erenstein O, Jaleta M, Sonder K, Mottaleb K, Prasanna BM. Global maize production, consumption and trade: trends and R&D implications. Food Secur. 2022;:1295–319. The Global Action for Fall Armyworm Control: Action framework 2020–2022. 2020. Prasanna BM, Nair SK, Babu R, Gowda M, Zhang X, Xu Y, et al. Increasing genetic gains in maize in stress-prone environments of the tropics. In: Genomic Designing of Climate-Smart Cereal Crops. 2020. Tarekegne A, Wegary D, Cairns JE, Zaman-Allah M, Beyene Y, Negera D, et al. Genetic gains in early maturing maize hybrids developed by the International Maize and Wheat Improvement Center in Southern Africa during 2000–2018. Front Plant Sci. 2023;14:1321308. Ndlovu N, Spillane C, McKeown PC, Cairns JE, Das B, Gowda M. Genome-wide association studies of grain yield and quality traits under optimum and low-nitrogen stress in tropical maize (Zea mays L.). Theor Appl Genet. 2022;135:4351–70. Raza A, Mubarik MS, Sharif R, Habib M, Jabeen W, Zhang C, et al. Developing drought-smart, ready-to-grow future crops. Plant Genome. 2023;16:e20279. Yuan Y, Cairns JE, Babu R, Gowda M, Makumbi D, Magorokosho C, et al. Genome-wide association mapping and genomic prediction analyses reveal the genetic architecture of grain yield and flowering time under drought and heat stress conditions in maize. Front Plant Sci. 2019;9 January:1–15. Masuka B, Magorokosho C, Olsen M, Atlin GN, Bänziger M, Pixley K V., et al. Gains in maize genetic improvement in eastern and southern Africa: II. CIMMYT open-pollinated variety breeding pipeline. Crop Sci. 2017;57:180–91. Fisher M, Abate T, Lunduka RW, Asnake W, Alemayehu Y, Madulu RB. Drought tolerant maize for farmer adaptation to drought in sub-Saharan Africa: Determinants of adoption in eastern and southern Africa. Clim Change. 2015;133:283–99. Prasanna BM, Cairns JE, Zaidi PH, Beyene Y, Makumbi D, Gowda M, et al. Beat the stress: breeding for climate resilience in maize for the tropical rainfed environments. Theor Appl Genet. 2021;134:1729–52. Brouwer, C. and Heibloem M. Irrigation Water Management: Irrigation Water Needs. Training Manual. Train Man No 3. 1980. He Z, Zhang T, Liu X, Shang X. Water-yield relationship responses of maize to ridge-furrow planting systems coupled with multiple irrigation levels in China’s Horqin Sandy land. Agronomy. 2018;8. Badu-Apraku B, Fakorede MAB. Breeding Maize for Drought Tolerance. Springer, Cham; 2017. Menkir A, Crossa J, Meseka S, Bossey B, Muhyideen O, Riberio PF, et al. Stacking Tolerance to Drought and Resistance to a Parasitic Weed in Tropical Hybrid Maize for Enhancing Resilience to Stress Combinations. Front Plant Sci. 2020;11:166. Ao S, Russelle MP, Varga T, Feyereisen GW, Coulter JA. Drought tolerance in maize is influenced by timing of drought stress initiation. Crop Sci. 2020;60:1591–606. Edmeades, G. O., Trevisan, W., Prasanna, B. M., & Campos H. Tropical Maize (Zea mays L.).In:Genetic Improvement of Tropical Crops. Springer Cham; 2017. Menkir A, Dieng I, Gedil M, Mengesha W, Oyekunle M, Riberio PF, et al. Approaches and progress in breeding drought-tolerant maize hybrids for tropical lowlands in west and central Africa. Plant Genome. 2024;17. McMillen MS, Mahama AA, Sibiya J, Lübberstedt T, Suza WP. Improving drought tolerance in maize: Tools and techniques. Frontiers in Genetics. 2022;13 October:1–13. Bänziger M, Edmeades GO, Beck D, Bellon M. Breeding for Drought and Nitrogen Stress Tolerance in Maize: From Theory to Practice. Mex DF CIMMYT. 2000;:68. Cooper M, Gho C, Leafgren R, Tang T, Messina C. Breeding drought-tolerant maize hybrids for the US corn-belt: Discovery to product. Journal of Experimental Botany. 2014;65:6191–4. Cooper M, Messina CD. Breeding crops for drought-Affected environments and improved climate resilience. Plant Cell. 2023;35:162–86. Wang N, Liu B, Liang X, Zhou Y, Song J, Yang J, et al. Genome-wide association study and genomic prediction analyses of drought stress tolerance in China in a collection of off-PVP maize inbred lines. Mol Breed. 2019;39. Varshney RK, Barmukh R, Bentley A, Nguyen HT. Exploring the genomics of abiotic stress tolerance and crop resilience to climate change. Plant Genome. 2024; February:1–6. Li C, Sun B, Li Y, Liu C, Wu X, Zhang D, et al. Numerous genetic loci identified for drought tolerance in the maize nested association mapping populations. BMC Genomics. 2016;17:1–11. Liu S, Qin F. Genetic dissection of maize drought tolerance for trait improvement. In: Molecular Breeding. 2021. p. 8. Shi W, Hao C, Zhang Y, Cheng J, Zhang Z, Liu J, et al. A combined association mapping and linkage analysis of kernel number per spike in common wheat (Triticum aestivum L.). Front Plant Sci. 2017;8 August:1–13. Davoud Torkamaneh FB. Genome-Wide Association Studies. 1st edition. New York, NY: Humana New York, NY; 2022. Alseekh S, Kostova D, Bulut M, Fernie AR. Genome-wide association studies: assessing trait characteristics in model and crop plants. Cellular and Molecular Life Sciences. 2021;78:5743–54. Xiao Y, Liu H, Wu L, Warburton M, Yan J. Genome-wide Association Studies in Maize: Praise and Stargaze. Mol Plant. 2017;10:359–74. Tibbs Cortes L, Zhang Z, Yu J. Status and prospects of genome-wide association studies in plants. Plant Genome. 2021;14. Hao D, Xue L, Zhang Z, Cheng Y, Chen G, Zhou G, et al. Combined linkage and association mapping reveal candidate loci for kernel size and weight in maize. Breed Sci. 2019;69:420–8. Zheng H, Chen J, Mu C, Makumbi D, Xu Y, Mahuku G. Combined linkage and association mapping reveal QTL for host plant resistance to common rust (Puccinia sorghi) in tropical maize. BMC Plant Biol. 2018;18:1–14. An Y, Chen L, Li YX, Li C, Shi Y, Zhang D, et al. Genome-wide association studies and whole-genome prediction reveal the genetic architecture of KRN in maize. BMC Plant Biol. 2020;20:1–11. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38:904–9. Zhang Z, Ersoz E, Lai CQ, Todhunter RJ, Tiwari HK, Gore MA, et al. Mixed linear model approach adapted for genome-wide association studies. Nat Genet. 2010;42:355–60. Li M, Liu X, Bradbury P, Yu J, Zhang YM, Todhunter RJ, et al. Enrichment of statistical power for genome-wide association studies. BMC Biol. 2014;12. Kang HM, Sul JH, Service SK, Zaitlen NA, Kong SY, Freimer NB, et al. Variance component model to account for sample structure in genome-wide association studies. Nat Genet. 2010;42:348–54. Zhou X, Stephens M. Genome-wide efficient mixed-model analysis for association studies. Nat Genet. 2012;44:821–4. Yoosefzadeh-Najafabadi M, Eskandari M, Belzile F, Torkamaneh D. Genome-Wide Association Study Statistical Models: A Review. In: Methods in Molecular Biology. 2022. p. 43–62. Zhang YW, Tamba CL, Wen YJ, Li P, Ren WL, Ni YL, et al. mrMLM v4.0.2: An R Platform for Multi-locus Genome-wide Association Studies. Genomics, Proteomics Bioinforma. 2020;18:481–7. Zhang YM, Jia Z, Dunwell JM. Editorial: The applications of new multi-locus GWAS methodologies in the genetic dissection of complex traits, volume II. Front Plant Sci. 2023;14 December:1–3. Liu, X., Huang, M., Fan, B., Buckler, E. S., & Zhang Z. Iterative Usage of Fixed and Random Effect Models for Powerful and Efficient Genome- Wide Association Studies. PLOS Genet. 2016;12:1–24. Huang M, Liu X, Zhou Y, Summers RM, Zhang Z. BLINK: A package for the next level of genome-wide association studies with both individuals and markers in the millions. Gigascience. 2019;8:1–12. Wang SB, Feng JY, Ren WL, Huang B, Zhou L, Wen YJ, et al. Improving power and accuracy of genome-wide association studies via a multi-locus mixed linear model methodology. Sci Rep. 2016;6. Tamba CL, Zhang Y-M. A fast mrMLM algorithm for multi-locus genome-wide association studies. bioRxiv. 2018. Wen YJ, Zhang H, Ni YL, Huang B, Zhang J, Feng JY, et al. Methodological implementation of mixed linear models in multi-locus genome-wide association studies. Brief Bioinform. 2018;19:700–12. Zhang J, Feng JY, Ni YL, Wen YJ, Niu Y, Tamba CL, et al. PLARmEB: Integration of least angle regression with empirical Bayes for multilocus genome-wide association studies. Heredity (Edinb). 2017;118:517–24. Ren WL, Wen YJ, Dunwell JM, Zhang YM. PKWmEB: Integration of Kruskal-Wallis test with empirical Bayes under polygenic background control for multi-locus genome-wide association study. Heredity (Edinb). 2018;120:208–18. Tamba CL, Ni YL, Zhang YM. Iterative sure independence screening EM-Bayesian LASSO algorithm for multi-locus genome-wide association studies. PLoS Comput Biol. 2017;13. Khan N, Shazadee H, Cloutier S, You FM. Genomics Assisted Breeding Strategy in Flax. In: You FM, Fofana B, editors. The Flax Genome. Compendium of Plant Genomes. 1st edition. Springer, Cham; 2023. p. 253–72. Meuwissen THE, Hayes BJ, Goddard ME. Prediction of total genetic value using genome-wide dense marker maps. Genetics. 2001;157:1819–29. Crossa J, Pérez-Rodríguez P, Cuevas J, Montesinos-López O, Jarquín D, de los Campos G, et al. Genomic Selection in Plant Breeding: Methods, Models, and Perspectives. Trends in Plant Science. 2017;22:961–75. Beyene Y, Semagn K, Mugo S, Tarekegne A, Babu R, Meisel B, et al. Genetic gains in grain yield through genomic selection in eight bi-parental maize populations under drought stress. Crop Sci. 2015;55:154–63. Wallace JG, Zhang X, Beyene Y, Semagn K, Olsen M, Prasanna BM, et al. Genome-wide association for plant height and flowering time across 15 tropical maize populations under managed drought stress and well-watered conditions in Sub-Saharan Africa. Crop Sci. 2016;56:2365–78. Zhang A, Wang H, Beyene Y, Semagn K, Liu Y, Cao S, et al. Effect of trait heritability, training population size and marker density on genomic prediction accuracy estimation in 22 bi-parental tropical maize populations. Front Plant Sci. 2017;8:1916. Zhang X, Pérez-Rodríguez P, Semagn K, Beyene Y, Babu R, López-Cruz MA, et al. Genomic prediction in biparental tropical maize populations in water-stressed and well-watered environments using low-density and GBS SNPs. Heredity (Edinb). 2015;114:291–9. Beyene Y, Gowda M, Pérez-Rodríguez P, Olsen M, Robbins KR, Burgueño J, et al. Application of Genomic Selection at the Early Stage of Breeding Pipeline in Tropical Maize. Front Plant Sci. 2021;12:1265. Masuka B, Araus JL, Das B, Sonder K, Cairns JE. Phenotyping for Abiotic Stress Tolerance in Maize. Journal of Integrative Plant Biology. 2012;54:238–49. Zaidi PH. Management of drought stress in field phenotyping. Mexico City Mexico: CIMMYT; 2019. Alvarado G, Rodríguez FM, Pacheco A, Burgueño J, Crossa J, Vargas M, et al. META-R: A software to analyze data from multi-environment plant breeding trials. Crop J. 2020;8:745–56. Bates D, Mächler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67. Kresovich S. Quantitative genetics in maize breeding. F Crop Res. 1990;23:78–9. Semagn K. Leaf tissue sampling and DNA extraction protocols. Methods Mol Biol. 2014;1115:53–67. Elshire RJ, Glaubitz JC, Sun Q, Poland JA, Kawamoto K, Buckler ES, et al. A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS One. 2011;6:19379. Glaubitz JC, Casstevens TM, Lu F, Harriman J, Elshire RJ, Sun Q, et al. TASSEL-GBS: A high capacity genotyping by sequencing analysis pipeline. PLoS One. 2014;9. Bradbury PJ, Zhang Z, Kroon DE, Casstevens TM, Ramdoss Y, Buckler ES. TASSEL: Software for association mapping of complex traits in diverse samples. Bioinformatics. 2007;23:2633–5. Hubisz MJ, Falush D, Stephens M, Pritchard JK. Inferring weak population structure with the assistance of sample group information. Mol Ecol Resour. 2009;9:1322–32. Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155:945–59. Earl DA, vonHoldt BM. STRUCTURE HARVESTER: A website and program for visualizing STRUCTURE output and implementing the Evanno method. Conserv Genet Resour. 2012;4:359–61. Evanno G, Regnaut S, Goudet J. Detecting the number of clusters of individuals using the software STRUCTURE: A simulation study. Mol Ecol. 2005;14:2611–20. Wang J, Zhang Z. GAPIT Version 3: Boosting Power and Accuracy for Genomic Association and Prediction. Genomics, Proteomics Bioinforma. 2021;19:629–40. Lipka AE, Tian F, Wang Q, Peiffer J, Li M, Bradbury PJ, et al. GAPIT: Genome association and prediction integrated tool. Bioinformatics. 2012;28:2397–9. VanRaden PM. Efficient methods to compute genomic predictions. J Dairy Sci. 2008;91:4414–23. Hill WG, Weir BS. Variances and covariances of squared linkage disequilibria in finite populations. Theor Popul Biol. 1988;33:54–78. Remington DL, Thornsberry JM, Matsuoka Y, Wilson LM, Whitt SR, Doebley J, et al. Structure of linkage disequilibrium and phenotypic associations in the maize genome. Proc Natl Acad Sci U S A. 2001;98:11479–84. Li MX, Yeung JMY, Cherny SS, Sham PC. Evaluating the effective numbers of independent tests and significant p-value thresholds in commercial genotyping arrays and public imputation reference datasets. Hum Genet. 2012;131:747–56. Gao X, Starmer ÃJ, Martin ER. A Multiple Testing Correction Method for Genetic Association Studies Using Correlated Single Nucleotide Polymorphisms. Genet Epedemiology. 2008;369:361–9. Slaten ML, Chan YO, Shrestha V, Lipka AE, Angelovici R. HAPPI GWAS: Holistic analysis with pre- And post-integration GWAS. Bioinformatics. 2020;36:4655–7. Jiao Y, Peluso P, Shi J, Liang T, Stitzer MC, Wang B, et al. Improved maize reference genome with single-molecule technologies Complete and accurate reference genomes and annotations provide fundamental tools for characterization of genetic and functional variation. Nat Publ Gr. 2017. Schnable PS, Ware D, Fulton RS, Stein JC, Wei F, Pasternak S, et al. The B73 maize genome: Complexity, diversity, and dynamics. Science (80- ). 2009;326:1112–5. Danecek P, Auton A, Abecasis G, Albers CA, Banks E, DePristo MA, et al. The variant call format and VCFtools. Bioinformatics. 2011;27:2156–8. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: Analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263–5. Zhang R, Jia G, Diao X. geneHapR: an R package for gene haplotypic statistics and visualization. BMC Bioinformatics. 2023;24. Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, et al. The structure of haplotype blocks in the human genome. Science (80- ). 2002;296:2225–9. Endelman JB. Ridge Regression and Other Kernels for Genomic Selection with R Package rrBLUP. Plant Genome. 2011;4:250–5. Wickham H. ggplot2. Springer-V. New York, NY: Springer New York; 2009. Cairns JE, Prasanna BM. Developing and deploying climate-resilient maize varieties in the developing world. Curr Opin Plant Biol. 2018;45:226–30. Wang Y, Guo H, Wu X, Wang J, Li H, Zhang R. Transcriptomic and physiological responses of contrasting maize genotypes to drought stress. Front Plant Sci. 2022;13. Blancon J, Buet C, Dubreuil P, Tixier MH, Baret F, Praud S. Maize green leaf area index dynamics: genetic basis of a new secondary trait for grain yield in optimal and drought conditions. Theor Appl Genet. 2024;137:68. Cairns JE, Crossa J, Zaidi PH, Grudloyma P, Sanchez C, Luis Araus J, et al. Identification of drought, heat, and combined drought and heat tolerant donors in maize. Crop Sci. 2013;53:1335–46. Trachsel S, Leyva M, Lopez M, Suarez EA, Mendoza A, Montiel NG, et al. Identification of tropical maize germplasm with tolerance to drought, nitrogen deficiency, and combined heat and drought stresses. Crop Sci. 2016;56. Rezende WS, Beyene Y, Mugo S, Ndou E, Gowda M, Sserumaga JP, et al. Performance and yield stability of maize hybrids in stress-prone environments in eastern Africa. Crop J. 2020;8:107–18. Beyene Y, Mugo S, Semagn K, Asea G, Trevisan W, Tarekegne A, et al. Genetic distance among doubled haploid maize lines and their testcross performance under drought stress and non-stress conditions. Euphytica. 2013;192:379–92. Chapman SC, Edmeades GO. Selection improves drought tolerance in tropical maize populations: II. Direct and correlated responses among secondary traits. Crop Sci. 1999;39:1315–24. Ndlovu N, Gowda M, Beyene Y, Chaikam V, Nzuve FM, Makumbi D, et al. Genomic loci associated with grain yield under well-watered and water-stressed conditions in multiple bi-parental maize populations. Front Sustain Food Syst. 2024;8. Bolaños J, Edmeades GO. The importance of the anthesis-silking interval in breeding for drought tolerance in tropical maize. F Crop Res. 1996;48. Badu-Apraku B, Akinwale RO, Franco J, Oyekunle M. Assessment of reliability of secondary traits in selecting for improved grain yield in drought and low-nitrogen environments. Crop Sci. 2012;52:2050–62. Zhang X, Ren Z, Luo B, Zhong H, Ma P, Zhang H, et al. Genetic architecture of maize yield traits dissected by QTL mapping and GWAS in maize. Crop J. 2022;10:436–46. Clark RM, Linton E, Messing J, Doebley JF. Pattern of diversity in the genomic region near the maize domestication gene tb1. Proc Natl Acad Sci U S A. 2004;101:700–7. Rashid Z, Sofi M, Harlapur SI, Kachapur RM, Dar ZA, Singh PK, et al. Genome-wide association studies in tropical maize germplasm reveal novel and known genomic regions for resistance to Northern corn leaf blight. Sci Rep. 2020;10:1–16. Lu Y, Shah T, Hao Z, Taba S, Zhang S, Gao S, et al. Comparative SNP and haplotype analysis reveals a higher genetic diversity and rapider LD decay in tropical than temperate germplasm in maize. PLoS One. 2011;6. Xiong X, Li J, Su P, Duan H, Sun L, Xu S, et al. Genetic dissection of maize (Zea mays L.) chlorophyll content using multi-locus genome-wide association studies. BMC Genomics. 2023;24. Zhou G, Zhu Q, Mao Y, Chen G, Xue L, Lu H, et al. Multi-Locus Genome-Wide Association Study and Genomic Selection of Kernel Moisture Content at the Harvest Stage in Maize. Front Plant Sci. 2021;12 July:1–13. Wen YJ, Wu X, Wang S, Han L, Shen B, Wang Y, et al. Identification of QTN-by-environment interactions for yield related traits in maize under multiple abiotic stresses. Front Plant Sci. 2023;14 February. Grotewold E, Gray J. Maize transcription factors. Handb Maize Genet Genomics. 2009;II:693–713. Ningning Z, Binbin L, Fan Y, Jianzhong C, Yuqian Z, Yejian W, et al. Molecular mechanisms of drought resistance using genome-wide association mapping in maize (Zea mays L.). BMC Plant Biol. 2023;23:1–17. Zhu Y, Liu Y, Zhou K, Tian C, Aslam M, Zhang B, et al. Overexpression of ZmEREBP60 enhances drought tolerance in maize. J Plant Physiol. 2022;275. Qi H, Liang K, Ke Y, Wang J, Yang P, Yu F, et al. Advances of Apetala2/Ethylene Response Factors in Regulating Development and Stress Response in Maize. International Journal of Molecular Sciences. 2023;24. Zhang J, Liao J, Ling Q, Xi Y, Qian Y. Genome-wide identification and expression profiling analysis of maize AP2/ERF superfamily genes reveal essential roles in abiotic stress tolerance. BMC Genomics. 2022;23. Farmer LM, Book AJ, Lee KH, Lin YL, Fu H, Vierstraa RD. The RAD23 family provides an essential connection between the 26S proteasome and ubiquitylated proteins in Arabidopsis. Plant Cell. 2010;22:124–42. Doroodian P, Hua Z. The ubiquitin switch in plant stress response. Plants. 2021;10:1–21. Wang X, An Y, Xu P, Xiao J. Functioning of PPR Proteins in Organelle RNA Metabolism and Chloroplast Biogenesis. Frontiers in Plant Science. 2021;12 February:1–8. Xing H, Fu X, Yang C, Tang X, Guo L, Li C, et al. Genome-wide investigation of pentatricopeptide repeat gene family in poplar and their expression analysis in response to biotic and abiotic stresses. Sci Rep. 2018;8:1–9. Zhu X, Ze M, Yin J, Chern M, Wang M, Zhang X, et al. A phosphofructokinase B-type carbohydrate kinase family protein, PFKB1, is essential for chloroplast development at early seedling stage in rice. Plant Sci. 2020;290. Peng Li , Zhaoxia Li GX and JZ*. Trihelix Transcription Factor ZmThx20 Is Required for Kernel Development in Maize. Int J Mol Sci. 2021;22:12137. Zhao D, Gao F, Guan P, Gao J, Guo Z, Guo J, et al. Identification and analysis of differentially expressed trihelix genes in maize (Zea mays) under abiotic stresses. PeerJ. 2023;11. Riggs JW, Rockwell NC, Cavales PC, Callis J. Identification of the plant ribokinase and discovery of a role for Arabidopsis Ribokinase in nucleoside metabolism. J Biol Chem. 2016;291:22572–82. Yang J, Tang Z, Yang W, Huang Q, Wang Y, Huang M, et al. Genome-wide characterization and identification of Trihelix transcription factors and expression profiling in response to abiotic stresses in Chinese Willow (Salix matsudana Koidz). Front Plant Sci. 2023;14. Ran F, Wang Y, Jiang F, Yin X, Bi Y, Shaw RK, et al. Studies on Candidate Genes Related to Flowering Time in a Multiparent Population of Maize Derived from Tropical and Temperate Germplasm. Plants. 2024;13. Khan SU, Zheng Y, Chachar Z, Zhang X, Zhou G, Zong N, et al. Dissection of Maize Drought Tolerance at the Flowering Stage Using Genome-Wide Association Studies. Genes (Basel). 2022;13. Lu Y, Zhang S, Shah T, Xie C, Hao Z, Li X, et al. Joint linkage-linkage disequilibrium mapping is a powerful approach to detecting quantitative trait loci underlying drought tolerance in maize. Proc Natl Acad Sci U S A. 2010;107:19585–90. Raunser S, Magnani R, Huang Z, Houtz RL, Trievel RC, Penczek PA, et al. Rubisco in complex with Rubisco large subunit methyltransferase. Proc Natl Acad Sci U S A. 2009;106:3160–5. Mininno M, Brugière S, Pautre V, Gilgen A, Ma S, Ferro M, et al. Characterization of chloroplastic fructose 1,6-bisphosphate aldolases as lysine-methylated proteins in plants. J Biol Chem. 2012;287:21034–44. Li J, Zhang Y, Gao Z, Xu X, Wang Y, Lin Y, et al. Plant U-box E3 ligases PUB25 and PUB26 control organ growth in Arabidopsis. New Phytol. 2021;229:403–13. Trujillo M. News from the PUB: Plant U-box type E3 ubiquitin ligases. Journal of Experimental Botany. 2018;69:371–84. Liu Y, Li C, Qin A, Deng W, Chen R, Yu H, et al. Genome-wide identification and transcriptome profiling expression analysis of the U-box E3 ubiquitin ligase gene family related to abiotic stress in maize (Zea mays L.). BMC Genomics. 2024;25. Jia T, Ge Q, Zhang S, Zhang Z, Liu A, Fan S, et al. UDP-Glucose Dehydrogenases: Identification, Expression, and Function Analyses in Upland Cotton (Gossypium hirsutum). Front Genet. 2021;11. Kärkönen A, Murigneux A, Martinant JP, Pepey E, Tatout C, Dudley BJ, et al. UDP-glucose dehydrogenases of maize: A role in cell wall pentose biosynthesis. Biochem J. 2005;391:409–15. Wu H, Bai B, Lu X, Li H. A gibberellin-deficient maize mutant exhibits altered plant height, stem strength and drought tolerance. Plant Cell Rep. 2023;42:1687–99. Martín-Trillo M, Cubas P. TCP genes: a family snapshot ten years later. Trends in Plant Science. 2010;15:31–9. Lan T, Xiong W, Chen X, Mo B, Tang G. Plant cytoplasmic ribosomal proteins: an update on classification, nomenclature, evolution and resources. Plant J. 2022;110:292–318. Li Y, Liu F, Li P, Wang T, Zheng C, Hou B. An Arabidopsis Cytokinin-Modifying Glycosyltransferase UGT76C2 Improves Drought and Salt Tolerance in Rice. Front Plant Sci. 2020;11 November:1–15. Jia H, Li M, Li W, Liu L, Jian Y, Yang Z, et al. A serine/threonine protein kinase encoding gene KERNEL NUMBER PER ROW6 regulates maize grain yield. Nat Commun. 2020;11. Shohat H, Eliaz NI, Weiss D. Gibberellin in tomato: metabolism, signaling and role in drought responses. Molecular Horticulture. 2021;1. Li S. The Arabidopsis thaliana TCP transcription factors: A broadening horizon beyond development. Plant Signal Behav. 2015;10. Liu Y, Chen Z, Zhang C, Guo J, Liu Q, Yin Y, et al. Gene editing of ZmGA20ox3 improves plant architecture and drought tolerance in maize. Plant Cell Rep. 2024;43. Jiao P, Liu T, Zhao C, Fei J, Guan S, Ma Y. ZmTCP14, a TCP transcription factor, modulates drought stress response in Zea mays L. Environ Exp Bot. 2023;208. Ding S, Cai Z, Du H, Wang H. Genome-wide analysis of TCP family genes in Zea mays l. Identified a role for ZmTCP42 in drought tolerance. Int J Mol Sci. 2019;20. Takemiya A, Sugiyama N, Fujimoto H, Tsutsumi T, Yamauchi S, Hiyama A, et al. Phosphorylation of BLUS1 kinase by phototropins is a primary step in stomatal opening. Nat Commun. 2013;4. Zeiner A, Colina FJ, Citterico M, Wrzaczek M. CYSTEINE-RICH RECEPTOR-LIKE PROTEIN KINASES: their evolution, structure, and roles in stress response and development. Journal of Experimental Botany. 2023;74:4910–27. Zhang Y, Tian H, Chen D, Zhang H, Sun M, Chen S, et al. Cysteine-rich receptor-like protein kinases: emerging regulators of plant stress responses. Trends in Plant Science. 2023;28:776–94. Li C, Chen G, Mishina K, Yamaji N, Ma JF, Yukuhiro F, et al. A GDSL-motif esterase/acyltransferase/lipase is responsible for leaf water retention in barley. Plant Direct. 2017;1. Shen G, Sun W, Chen Z, Shi L, Hong J, Shi J. Plant GDSL Esterases/Lipases: Evolutionary, Physiological and Molecular Functions in Plant Development. Plants. 2022;11. Su HG, Zhang XH, Wang TT, Wei WL, Wang YX, Chen J, et al. Genome-Wide Identification, Evolution, and Expression of GDSL-Type Esterase/Lipase Gene Family in Soybean. Front Plant Sci. 2020;11. Li Z, Ahammed GJ. Plant stress response and adaptation via anthocyanins: A review. Plant Stress. 2023;10. Lu N, Jun JH, Li Y, Dixon RA. An unconventional proanthocyanidin pathway in maize. Nat Commun. 2023;14. Zhang Q, Liu H, Wu X, Wang W. Identification of drought tolerant mechanisms in a drought-tolerant maize mutant based on physiological, biochemical and transcriptomic analyses. BMC Plant Biol. 2020;20:1–14. Kushwaha HR, Joshi R, Pareek A, Singla-Pareek SL. MATH-domain family shows response toward abiotic stress in arabidopsis and rice. Front Plant Sci. 2016;7. Cai G, Zang Y, Wang Z, Liu S, Wang G. Arabidopsis BTB-A2s Play a Key Role in Drought Stress. Biology (Basel). 2024;13. Zeng T, Meng Z, Yue R, Lu S, Li W, Li W, et al. Genome wide association analysis for yield related traits in maize. BMC Plant Biol. 2022;22:1–11. Guo J, Sun B, He H, Zhang Y, Tian H, Wang B. Current understanding of bhlh transcription factors in plant abiotic stress tolerance. International Journal of Molecular Sciences. 2021;22. Zhang T, Lv W, Zhang H, Ma L, Li P, Ge L, et al. Genome-wide analysis of the basic Helix-Loop-Helix (bHLH) transcription factor family in maize. BMC Plant Biol. 2018;18. Nan GL, Teng C, Fernandes J, O’Connor L, Meyers BC, Walbot V. A cascade of bHLH-regulated pathways programs maize anther development. Plant Cell. 2022;34:1207–25. He Z, Zhong J, Sun X, Wang B, Terzaghi W, Dai M. The maize ABA receptors ZmPYl8, 9, and 12 facilitate plant drought resistance. Front Plant Sci. 2018;9. Niu L, Liu L, Wang W. Digging for Stress-Responsive Cell Wall Proteins for Developing Stress-Resistant Maize. Front Plant Sci. 2020;11. Song N, Xu Z, Wang J, Qin Q, Jiang H, Si W, et al. Genome-wide analysis of maize CONSTANS-LIKE gene family and expression profiling under light/dark and abscisic acid treatment. Gene. 2018;673:1–11. Wang F, Han T, Jeffrey Chen Z. Circadian and photoperiodic regulation of the vegetative to reproductive transition in plants. Commun Biol. 2024;7:1–11. Bhat JA, Yu D, Bohra A, Ganie SA, Varshney RK. Features and applications of haplotypes in crop breeding. Commun Biol. 2021;4:1–12. Beyene Y, Gowda M, Olsen M, Robbins KR, Pérez-Rodríguez P, Alvarado G, et al. Empirical Comparison of Tropical Maize Hybrids Selected Through Genomic and Phenotypic Selections. Front Plant Sci. 2019;10:1502. Ertiro BT, Labuschagne M, Olsen M, Das B, Prasanna BM, Gowda M. Genetic Dissection of Nitrogen Use Efficiency in Tropical Maize Through Genome-Wide Association and Genomic Prediction. Front Plant Sci. 2020;11 April:1–16. Tables Tables 1 to 2 are available in the Supplementary Files section Additional Declarations No competing interests reported. Supplementary Files SuplemenetaryTable.xlsx Supplementary Table S1. Pedigree of 236 lines used in the study. Supplementary Table S2. Summary of SNPs across the ten maize chromosomes. Supplementary Table S3. Significant QTNs identified for eight traits across multi-environment under optimum conditions using six multi-locus GWAS models. Supplementary Table S4| Significant QTNs identified for eight traits across multi-environment under drought conditions using six multi-locus GWAS models. Supplementary Table S5. Significant SNPs associated with eight traits under drought and optimum condition using Farm CPU GWAS model from GAPIT. Supplementary Table S6. Number of stable QTNs detected by at least two GWAS models for grain yield, flowering traits and other agronomic traits under drought and optimum conditions. SupplementaryFigureS1.jpg Supplementary Figure S1. Population structure of 236 maize inbred lines based on of 215,542 SNPs markers: (a) Evanno plot of the number of clusters (K) against delta K to determine the optimum number of K; (b) population structure of the 236 lines at K = 3 to K = 10. Each individual is shown as a vertical line divided into K colored segments, with segment lengths indicating the estimated probability of membership to each cluster. SupplementaryFigureS2.jpg Supplementary Figure S2. Farm CPU Manhattan and Q-Q plots of genome-wide association study (GWAS) on eight traits evaluated under optimum (_OPT) and drought (_DS) environmental conditions. The − log10(p) values on the Y-axis in the Manhattan plot represent grain yield (GY), Days to 50 % anthesis, (AD), Days to 50 % silking (SD), Anthesis-Silking Interval (ASI), Plant height (PH), Ear height (EH), Ear position (EPO) and Ear per plant (EPP) plotted against chromosome position on X-axis. The red and blue solid horizontal lines in the Manhattan plots represent the genome-wide (− log10 (p) =6.2). The quantile—quantile plots represent observed against the expected −log10 (p). SupplementaryFigureS3.jpg Supplementary Figure S3. Manhattan and Q-Q plots of genome-wide association study (GWAS) on eight traits evaluated under optimum (OPT) and drought (_DS) environmental conditions for six GWAS models. The pink dots above the threshold indicates significant QTNs identified by more than one ML-GWAS models, while green and blue dots above the threshold represent significant QTNs identified by a single ML-GWAS model . The black horizontal dashed line indicates the genome-wide significance threshold, corresponding to a −log10 (p) value of a LOD score ≥ 3.0 for ML-GWAS models. Tables.docx Cite Share Download PDF Status: Published Journal Publication published 01 Feb, 2025 Read the published version in BMC Plant Biology → Version 1 posted Editorial decision: Revision requested 25 Oct, 2024 Editor assigned by journal 24 Oct, 2024 Submission checks completed at journal 23 Oct, 2024 First submitted to journal 18 Oct, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5289238","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":370025923,"identity":"e5bf8b49-5bb4-4685-8eda-aaec6c3700c6","order_by":0,"name":"Manigben Kulai Amadu","email":"","orcid":"","institution":"CSIR-Savanna Agricultural Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Manigben","middleName":"Kulai","lastName":"Amadu","suffix":""},{"id":370025924,"identity":"170a58ca-3f56-4419-9684-fc8269976e9b","order_by":1,"name":"Yoseph Beyene","email":"","orcid":"","institution":"International Maize and Wheat Improvement Center","correspondingAuthor":false,"prefix":"","firstName":"Yoseph","middleName":"","lastName":"Beyene","suffix":""},{"id":370025925,"identity":"bf75cbcb-6114-4c58-a72f-42fe9bf69a14","order_by":2,"name":"Vijay Chaikam","email":"","orcid":"","institution":"International Maize and Wheat Improvement Center","correspondingAuthor":false,"prefix":"","firstName":"Vijay","middleName":"","lastName":"Chaikam","suffix":""},{"id":370025926,"identity":"c2a1a16f-8d61-43f2-9bf4-342994efdc62","order_by":3,"name":"Pangirayi B. Tongoona","email":"","orcid":"","institution":"University of Ghana","correspondingAuthor":false,"prefix":"","firstName":"Pangirayi","middleName":"B.","lastName":"Tongoona","suffix":""},{"id":370025927,"identity":"cc94eafe-a6c9-4122-85b3-186340cecdff","order_by":4,"name":"Eric Y. Danquah","email":"","orcid":"","institution":"University of Ghana","correspondingAuthor":false,"prefix":"","firstName":"Eric","middleName":"Y.","lastName":"Danquah","suffix":""},{"id":370025928,"identity":"df9dca12-6df6-4b09-a46e-8d129781cbaf","order_by":5,"name":"Beatrice E. Ifie","email":"","orcid":"","institution":"University of Ghana","correspondingAuthor":false,"prefix":"","firstName":"Beatrice","middleName":"E.","lastName":"Ifie","suffix":""},{"id":370025929,"identity":"00d52305-23df-499e-a704-5e4fff204be0","order_by":6,"name":"Juan Burgueno","email":"","orcid":"","institution":"International Maize and Wheat Improvement Center","correspondingAuthor":false,"prefix":"","firstName":"Juan","middleName":"","lastName":"Burgueno","suffix":""},{"id":370025930,"identity":"946c2540-5c58-45e5-b484-d8d3376500e9","order_by":7,"name":"Boddupalli M Prasanna","email":"","orcid":"","institution":"International Maize and Wheat Improvement Center","correspondingAuthor":false,"prefix":"","firstName":"Boddupalli","middleName":"M","lastName":"Prasanna","suffix":""},{"id":370025931,"identity":"41bc8c8c-ac96-4e54-affb-dff3110a535b","order_by":8,"name":"Manje Gowda","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAyklEQVRIiWNgGAWjYNACNgY5BgbGBtK0GJOuJZF49boN7Bcf85TZpff3L278XMFgky/vQECL2QGeYmOec8m5M248bJY8w5BmufEAYS1p0rltzLkbJA42SDYwHDYwJOREqJb6dAOJg80/idTCfgyo5XCCAX9jG9gWeQI6GMwO8zAb/zl33HDGDcY2ywaDNAMDglqOtz98OKOsWp6///jjmw0VNgbyhBzGwMwDNVYiAUgA2QYHCGlhYH8AofmhSgnbMgpGwSgYBSMNAAAs4D+fIqL5+QAAAABJRU5ErkJggg==","orcid":"","institution":"International Maize and Wheat Improvement Center","correspondingAuthor":true,"prefix":"","firstName":"Manje","middleName":"","lastName":"Gowda","suffix":""}],"badges":[],"createdAt":"2024-10-18 11:53:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5289238/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5289238/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12870-025-06135-3","type":"published","date":"2025-02-01T15:57:16+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":67455117,"identity":"be233f1a-c636-46f1-a0f5-009abb90f795","added_by":"auto","created_at":"2024-10-25 08:35:22","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":795449,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/34306bb35344809fba753b34.png"},{"id":67454837,"identity":"b82152c0-f79f-456b-b2cb-eb4f1ee50f4b","added_by":"auto","created_at":"2024-10-25 08:27:22","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":483972,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Fugire2.png","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/218470997cb014cc47b275f1.png"},{"id":67455118,"identity":"58e0b07f-1e28-44b2-b083-64b58883a9f1","added_by":"auto","created_at":"2024-10-25 08:35:22","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":4100513,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/1cc46dddf57a3479a3ea3b89.png"},{"id":67453963,"identity":"e296815a-b7b8-4c66-be2a-575194e951a8","added_by":"auto","created_at":"2024-10-25 08:19:22","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":992267,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/c3a13ea49fb7a41b327e8360.png"},{"id":67453966,"identity":"3b9a3565-80c8-4653-a682-4757c55bd518","added_by":"auto","created_at":"2024-10-25 08:19:22","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1169207,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/201b80a3a78e1b03eddf855d.png"},{"id":67453961,"identity":"cdf87d89-0cd1-4e57-8daa-c877b6cb5105","added_by":"auto","created_at":"2024-10-25 08:19:22","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2316417,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/738272aa53b14ff5b2a11a8a.png"},{"id":67454841,"identity":"b40b989b-2035-4d1b-9ce9-cead671fc62b","added_by":"auto","created_at":"2024-10-25 08:27:22","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":169196,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/c5df43b7b9b07ad5eb406956.png"},{"id":75351944,"identity":"2315dfe9-78e6-4613-a157-0c7cef52ea1a","added_by":"auto","created_at":"2025-02-03 16:12:51","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":11907889,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/02db88c1-07f3-4911-87a4-fff6863beedf.pdf"},{"id":67453958,"identity":"fae902b4-e716-4875-898f-89bd6878c142","added_by":"auto","created_at":"2024-10-25 08:19:22","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":84842,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Table S1\u003c/strong\u003e. Pedigree of 236 lines used in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Table S2\u003c/strong\u003e. Summary of SNPs across the ten maize chromosomes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Table S3\u003c/strong\u003e. Significant QTNs identified for eight traits across multi-environment under optimum conditions using six multi-locus GWAS models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Table S4\u003c/strong\u003e| Significant QTNs identified for eight traits across multi-environment under drought conditions using six multi-locus GWAS models.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Table S5. \u003c/strong\u003eSignificant SNPs associated with eight traits under drought and optimum condition using Farm CPU GWAS model from GAPIT.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary Table S6. \u003c/strong\u003eNumber of stable QTNs detected by at least two GWAS models for grain yield, flowering traits and other agronomic traits under drought and optimum conditions.\u003c/p\u003e","description":"","filename":"SuplemenetaryTable.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/ba6837f7d4cca9826d132ee0.xlsx"},{"id":67454839,"identity":"dc7f3513-3339-4180-a3f3-6612dd6500f6","added_by":"auto","created_at":"2024-10-25 08:27:22","extension":"jpg","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":180620,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure S1. \u003c/strong\u003ePopulation structure of 236 maize inbred lines based on of 215,542 SNPs markers: (a) Evanno plot of the number of clusters (K) against delta K to determine the optimum number of K; (b) population structure of the 236 lines at K = 3 to K = 10. Each individual is shown as a vertical line divided into K colored segments, with segment lengths indicating the estimated probability of membership to each cluster.\u003c/p\u003e","description":"","filename":"SupplementaryFigureS1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/0b1698460d2c59367dd6871f.jpg"},{"id":67453967,"identity":"87b1b277-d452-4031-af40-79f37ea11981","added_by":"auto","created_at":"2024-10-25 08:19:22","extension":"jpg","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":1007285,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure S2. \u003c/strong\u003eFarm CPU Manhattan and Q-Q plots of genome-wide association study (GWAS) on eight traits evaluated under optimum (_OPT) and drought (_DS) environmental conditions. The − log10(p) values on the Y-axis in the Manhattan plot represent grain yield (GY), Days to 50 % anthesis, (AD), Days to 50 % silking (SD), Anthesis-Silking Interval (ASI), Plant height (PH), Ear height (EH), Ear position (EPO) and Ear per plant (EPP) plotted against chromosome position on X-axis. The red and blue solid horizontal lines in the Manhattan plots represent the genome-wide (− log10 (p) =6.2). The quantile—quantile plots represent observed against the expected −log10 (p).\u003c/p\u003e","description":"","filename":"SupplementaryFigureS2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/db81831d58c578a2fa42347e.jpg"},{"id":67453969,"identity":"8198a154-f6f2-495a-9c0a-bbbbb3606f52","added_by":"auto","created_at":"2024-10-25 08:19:22","extension":"jpg","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":669978,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure S3. \u003c/strong\u003eManhattan and Q-Q plots of genome-wide association study (GWAS) on eight traits evaluated under optimum (OPT) and drought (_DS) environmental conditions for six GWAS models. The pink dots above the threshold indicates significant QTNs identified by more than one ML-GWAS models, while green and blue dots above the threshold represent significant QTNs identified by a single ML-GWAS model . The black horizontal dashed line indicates the genome-wide significance threshold, corresponding to a −log10 (p) value of a LOD score ≥ 3.0 for ML-GWAS models.\u003c/p\u003e","description":"","filename":"SupplementaryFigureS3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/ef022817675f500d2984b367.jpg"},{"id":67454842,"identity":"efb0b9b4-7190-478c-95e8-e0bc8169151f","added_by":"auto","created_at":"2024-10-25 08:27:22","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":28757,"visible":true,"origin":"","legend":"","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-5289238/v1/2c081ee3fd7d857e23f341fc.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Genome-Wide Association Mapping and Genomic Prediction Analyses Reveal the Genetic Architecture of Grain Yield and Agronomic Traits under Drought and Optimum Conditions in Maize","fulltext":[{"header":"Introduction","content":"\u003cp\u003eMaize (\u003cem\u003eZea mays\u003c/em\u003e L.) is an indispensable cereal crop in global agri-food systems[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. However, grain yield is stagnating due to unpredictable climate change and increase in negative impacts of drought on maize production and productivity in sub-Saharan Africa (SSA) [\u003cspan additionalcitationids=\"CR4 CR5\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Boosting maize grain yield potential and improving maize resilience to drought are key solutions proposed for mitigating the effects of drought and climate changes while minimizing farmer\u0026rsquo;s risk [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. In SSA, it is estimated that about 40% of the region's maize-growing area experiences intermittent drought stress, leading to sizable yield reductions ranging from 10\u0026ndash;25% [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Maize plants require at least 500 to 800 mm of water, which translates to 5.6\u0026ndash;6.7 mm per day. This largely depends on the maturity group, soil moisture, growth stage, and environmental conditions [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Water deficit below this range can lead to severe water stress in maize plants, particularly during the flowering stage. The maize plants respond to drought stress by rolling their leaves, reducing leaf area, and closing stomata, which affects photosynthetic activity and enzyme production. Drought stress coinciding with the flowering and grain-filling stages in maize causes a sizable yield reduction of up to 90% [\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eConsiderable efforts have been made to boost the grain yield potential and stress-resilience in maize through conventional breeding[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. However, genetic improvement of grain yield under drought stress through conventional breeding methods has proven to be challenging [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], primarily caused by the multigenic nature of traits, controlled by many loci, each contributing a small effect[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. Owing to its multigenic nature of inheritance and genotype x environment interactions, grain yield often presents a low heritability under drought conditions [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. For this reason, it is difficult to accurately estimate breeding values, which results in lower genetic gain per unit time and thereby constrains the development of drought-tolerant maize hybrids[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. This, therefore, emphasizes the need to complement conventional breeding methods with genomic-assisted breeding tools to accelerate the development of high yielding drought-tolerant maize cultivars thereby boosting productivity in stress-prone areas.\u003c/p\u003e \u003cp\u003eRecent advances in crop genomics and phenomics have increased our understanding of the physiological and genetic basis of complex traits such as drought tolerance and grain yield [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Linkage mapping based on biparental population is one of the most powerful tools extensively utilized to identify several quantitative trait loci (QTL) related to grain yield and secondary traits under drought stress [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. However, many constructed maps suffer from low resolution and low allele richness [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Contrary to linkage mapping, GWAS utilizes diverse natural populations, which eliminates the need for developing segregating populations, saving time and cost[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] and hence is the most preferred method [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Moreover, it can detect multi-allelic variation, rare, and small effect QTLs simultaneously [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e], providing high resolution by leveraging historical meiotic recombination events available in diverse natural populations[\u003cspan additionalcitationids=\"CR31\" citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Conversely, the results of GWAS can be influenced by the population structure, which can significantly interfere with the power of QTL detection[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. For this reason, several statistical models have been developed including single-locus GWAS models and multi-locus GWAS models [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. In single-locus GWAS models such as the general linear model (GLM) [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]; Mixed linear model (MLM)[\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]; Enriched compressed mixed linear model (ECMLM)[\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]; efficient mixed model association eXpedited (EMMAX) [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e] and genome-wide efficient mixed-model association (GEMMA)[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e], test marker-trait associations for significance by multiple testing one marker at a time. These models incorporate population structure (e.g. Principal component, Kinship matrices, etc.) as fixed covariates or a random polygenic effect to address the genetic relatedness among individuals in a diverse population [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e] and tend to detect major QTLs. However, single locus models are often prone to high false positive rates or Type 1 errors. Bonferroni correction is commonly employed to control false positive rate (FPR)[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. However, the use of Bonferroni correction has been proven to be too conservative such that true quantitative trait nucleotides (QTNs) may be missed out when considering SNPs in linkage disequilibrium (LD). Therefore, multi-locus GWAS models have been recommended for addressing multiple test corrections[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMulti-locus GWAS models do not require Bonferroni correction, have higher power for detection of both major and minor QTL effects, and have proven to be superior to single locus models in detecting small effect loci[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. For this reason, several multi-locus GWAS including the multiple loci mixed model (MLMM), fixed and random model circulating probability unification (FarmCPU)[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e], and Bayesian-information and linkage-disequilibrium iteratively nested keyway (BLINK)[\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. In addition, other key models are multi-locus random-SNP-effect mixed linear model (mrMLM)[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e], the fast multi-locus random-SNP-effect mixed linear model (FASTmrMLM) [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e], the fast multi-locus random-SNP-effect efficient mixed-model association (FASTmrEMMA) [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e], polygenic-background-control-based least angle regression plus empirical Bayes (pLARmEB)[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e], polygenic-background-control-based Kruskal\u0026ndash;Wallis test plus empirical Bayes (pKWmEB) [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e], and Iterative sure independence screening expectation maximization Bayesian least absolute shrinkage and selection operator (ISIS EM-BLASSO) [\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e] are commonly used now. The multi-locus GWAS methods involve a two-step process. In the first step, a single-dimensional genome scan is implemented using a less stringent critical value to identify putative QTLs. In the second step, all putative QTLs identified in the first step are subjected to further genome-wide scan to identify true QTNs using a logarithm of the odds (LOD) statistics to determine their significance[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eGenomic selection (GS) or genomic prediction (GP) is another efficient genomic-assisted breeding tool in which genome-wide markers are fed into a prediction model to predict the genomic estimated breeding values (GEBVs) of lines in a breeding population [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. It has enormous potential for improving drought tolerance in maize, as it allows for more accurate selection of complex traits such as grain yield under drought and reduces the breeding cycles by enhancing the genetic gain per unit time [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. Several studies have demonstrated the effectiveness of utilizing GS for improving grain yield and secondary traits in drought tolerance breeding in maize[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan additionalcitationids=\"CR54\" citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e]. Zhang et al.[\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e] reported low to medium prediction accuracy for grain yield (GY) and secondary traits in maize under drought stress. The results indicated that several factors influenced the prediction accuracy values, including the types of breeding populations, the size of the training population, the complex nature of traits, the marker densities, and the genotyping platforms. Therefore, the combined application of different GWAS models and GS can enhance the power of QTL detection and accelerate breeding to improve drought tolerance and assist in selecting superior genotypes under drought stress.\u003c/p\u003e \u003cp\u003eThe objectives of the study were to (1) identify significant QTNs and putative candidate genes for GY and secondary traits under optimum and drought stress (DS) using multi-locus GWAS models and (2) assess the potential of GS in improving GY and related traits under DS and optimum conditions. These results will further deepen our understanding of the genetic architecture of complex traits, especially for GY and drought tolerance, which is critical for making accurate selections and developing stress-resilient maize hybrids.\u003c/p\u003e"},{"header":"Material and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eGermplasm, Experimental design, and Phenotyping\u003c/h2\u003e \u003cp\u003eA panel of 236 maize inbred lines was assembled for this study. The lines were developed by International Maize and Wheat Improvement Centre\u0026rsquo;s (CIMMYT) Global Maize Program through conventional breeding and doubled haploid technology [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. These elite maize lines were crossed with a popular single cross hybrid tester (CML566 x CML395) and produced 236 test cross hybrids. All test cross hybrids plus six commercial hybrids as checks (DK8031, H513, PH3253, Pioneer 3253, and WH505) were evaluated under DS and optimum conditions (Manigben et al., in review). The optimum experiments were conducted under rainfed conditions which was augmented with irrigation to avoid DS at seven (7) locations, that is, Embu [0.48\u0026deg; S 37.47\u0026deg; E, 1159 masl], Kaguru [37.67\u0026deg;E,-0.08\u0026deg;S,1463masl], Kakamega [0.29\u0026deg;N, 34.77\u0026deg;E,1535 masl], Kiboko[37.72\u0026deg;E, 2.22\u0026deg;S, 975 masl], Kirinyaga [37.19`E, 0.34`S, 1282masl], Mwtapa [3.93\u0026deg; S 39.74\u0026deg; E, 30 masl] and Shikutsa [ 0.28\u0026deg; N, 34.75\u0026deg; E ,1561 masl]. The drought experiment was conducted at three (3) locations, namely Kiboko [37.72\u0026deg;E, 2.22\u0026deg;S, 975 masl], Homabay [0.52\u0026deg;S, 34.45\u0026deg;E, 1751 masl] and Mtwapa [3.93\u0026deg;S, 39.74\u0026deg;E, 30 masl] during the dry season. The drought experiment was conducted following the protocol established by CIMMYT [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. The drought experiments were irrigated once a week using a drip irrigation system until two weeks before the expected flowering date. Irrigation was withdrawn to maintain DS until harvest.\u003c/p\u003e \u003cp\u003eThe experiments were set up in a 5 x 49 alpha lattice design with two replications. The experimental unit was a two-row plot of 5m long with intra-row spacing of 0.25 m and inter-row spacing of 0.75 m. Two seeds per hill were seeded and later thinned to one plant per hill three weeks after seedling emergence to a final plant population of 53,333 plants/ha. Basal fertilizer application was carried out at planting using di-ammonium phosphate (D.A.P) fertilizer at the rate of 60 Kg N and 60 Kg P\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e5\u003c/sub\u003e per hectare. Six weeks after emergence, all experiments were top-dressed with Urea at the rate of 60 Kg N. All the experiments were kept weeds-free by manual weeding and herbicide control. Detailed information on the pedigree of the inbred lines used in this study is presented in Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003ePhenotypic data collection and analysis\u003c/h3\u003e\n\u003cp\u003eIn all the experiments, data were collected on days to 50% anthesis (AD) and days to 50% silking (SD) as the number of days from planting to the day, when half of the plants per plot had shed pollen and silks emerged, respectively. Anthesis-silking interval (ASI) was computed as the difference between SD and AD. Plant height (PH) and ear height (EH) were measured in centimeters from the base of the plant to the height of the first tassel branch and the node bearing the upper ear, respectively. Ear position (EPO) was measured as the ratio of ear height to plant height per plot. The number of ears per plant (EPP) was determined by dividing the total number of ears per plot by the number of plants harvested per plot. Ears from each plot were shelled and weighed to determine grain yield (GY) in kilograms, which was then converted to tons per hectare (t/ha). Moisture content (MOI) of the shelled grains at harvest was measured with a portable handheld moisture meter and recorded in percentage. GY per plot in tons per hectare will be calculated using the field weight of harvested ears per plot and adjusted 12.5% moisture content. All trait measurements were done according to the procedure outlined in the drought phenotyping protocol of CIMMYT [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eData was analyzed for each and across locations under optimum and DS conditions. The restricted maximum likelihood (REML) estimates of variance components, coefficient of variation, broad-sense heritability, phenotypic and genetic correlation among traits for individual and combined analysis (Eq.\u0026nbsp;\u003cspan refid=\"Equ1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) were estimated using multi-environment trial analysis R package (META-R) [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. The linear mixed models available in META-R were implemented using the \u003cem\u003eLme4\u003c/em\u003e R-package [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e]. In the model, all factors were treated as random effects in this analysis except the genotype effect to estimate the best linear unbiased estimates (BLUEs). The best linear unbiased predictions (BLUPs) and variance components were estimated by treating all factors as random except replication and environments.\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{\\varvec{Y}}_{\\varvec{i}\\varvec{j}\\varvec{k}\\varvec{l}}=\\varvec{\\mu\\:}+\\:{\\varvec{E}\\varvec{n}\\varvec{v}}_{\\varvec{i}}+{\\varvec{R}\\varvec{e}\\varvec{p}}_{\\varvec{j}}\\left({\\varvec{E}\\varvec{n}\\varvec{v}}_{\\varvec{i}}\\right)+{\\varvec{B}\\varvec{l}\\varvec{o}\\varvec{c}\\varvec{k}}_{\\varvec{k}}\\left({\\varvec{R}\\varvec{e}\\varvec{p}}_{\\varvec{j}}{\\varvec{E}\\varvec{n}\\varvec{v}}_{\\varvec{i}}\\right)+{\\varvec{G}\\varvec{e}\\varvec{n}}_{\\varvec{l}}+\\:{\\varvec{E}\\varvec{n}\\varvec{v}}_{\\varvec{i}}\\:x\\:{\\varvec{G}\\varvec{e}\\varvec{n}}_{\\varvec{l}}+\\:+{\\varvec{\\epsilon\\:}}_{\\varvec{i}\\varvec{j}\\varvec{k}\\varvec{l}}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{\\varvec{Y}}_{\\varvec{i}\\varvec{j}\\varvec{k}\\varvec{l}}\$\u003c/span\u003e\u003c/span\u003e is the trait of interest, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:\\varvec{\\mu\\:}\$\u003c/span\u003e\u003c/span\u003e is the overall mean effect; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{\\varvec{E}\\varvec{n}\\varvec{v}}_{\\varvec{i}}\$\u003c/span\u003e\u003c/span\u003e is the effect of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{i}^{th}\$\u003c/span\u003e\u003c/span\u003eenvironment;\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{\\varvec{R}\\varvec{e}\\varvec{p}}_{\\varvec{j}}\\left({\\varvec{E}\\varvec{n}\\varvec{v}}_{\\varvec{i}}\\right)\$\u003c/span\u003e\u003c/span\u003e is the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{j}^{th}\$\u003c/span\u003e\u003c/span\u003e replication within \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{i}^{th}\$\u003c/span\u003e\u003c/span\u003e environment; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{\\varvec{B}\\varvec{l}\\varvec{o}\\varvec{c}\\varvec{k}}_{\\varvec{k}}\\left({\\varvec{R}\\varvec{e}\\varvec{p}}_{\\varvec{j}}{\\varvec{E}\\varvec{n}\\varvec{v}}_{\\varvec{i}}\\right)\$\u003c/span\u003e\u003c/span\u003e is the effect of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{k}^{th}\$\u003c/span\u003e\u003c/span\u003e incomplete block within\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:\\:{j}^{th}\$\u003c/span\u003e\u003c/span\u003ereplication in \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{i}^{th}\$\u003c/span\u003e\u003c/span\u003eenvironment; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{\\varvec{G}\\varvec{e}\\varvec{n}}_{\\varvec{l}}\$\u003c/span\u003e\u003c/span\u003e is effect of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{l}^{th}\$\u003c/span\u003e\u003c/span\u003egenotype; \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{\\varvec{E}\\varvec{n}\\varvec{v}}_{\\varvec{i}}\\:x\\:{\\varvec{G}\\varvec{e}\\varvec{n}}_{\\varvec{l}}\$\u003c/span\u003e\u003c/span\u003e is the effect of genotype x environment interactions and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\$\\:{\\varvec{\\epsilon\\:}}_{\\varvec{i}\\varvec{j}\\varvec{k}\\varvec{l}}\$\u003c/span\u003e\u003c/span\u003e is the residual effect. The variance components from the combined analysis were used to compute broad sense heritability [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003cb\u003eDNA extraction, Sequencing, SNP discovery, and calling.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eGenomic Deoxyribonucleic acid (DNA) of 236 lines was extracted from seedlings at the 4-leaf stage using a modified version of CIMMYT\u0026rsquo;s high throughput mini-prep Cetyl Trimethyl Ammonium Bromide (CTAB) protocol [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e]. The DNA samples were shipped to Cornell University for genotyping. In brief, the high-quality DNA extracted from each leaf sample of the 236 lines was digested with restriction endonuclease \u003cem\u003eApe KI.\u003c/em\u003e DNA libraries were constructed for each sample and sequenced using genotyping-by-sequencing (GBS) protocol as described by Elshire et al. [\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]. Raw GBS data of 955,690 SNPs distributed across the ten maize chromosomes were received from the Institute of Biotechnology at Cornell University, USA, after mapping to B73 AGPv2 coordinates. SNP calling was carried out using the TASSEL-GBS pipeline [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e]. The raw GBS data were cleaned by removal of SNP markers with a minimum count of 90%, greater than 5% heterozygosity, and less than 5% minor allele frequency using TASSEL software version 5.2 [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e], resulting in a total of 230,743 SNPs. In addition, the lines with greater than 20% missing data and SNPs not located on any of the ten chromosomes were further filtered out to a final dataset of 215,542 SNPs in 236 diverse lines for further analysis. The density and distribution map of SNPs on each of the ten maize chromosomes was drawn using a \u003cem\u003eCMplot\u003c/em\u003e R package (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/YinLiLin/CMplot\u003c/span\u003e\u003cspan address=\"https://github.com/YinLiLin/CMplot\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003ePopulation structure and linkage disequilibrium analysis\u003c/h3\u003e\n\u003cp\u003eTo capture population structure and cryptic genetic relatedness among the 236 lines, population structure, and kinship analysis were carried out using 215,542 genome-wide SNPs distributed across the ten chromosomes. The population structure was estimated using the admixture model method implemented in the software package STRUCTURE version 2.3.4 [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e]. The number of subpopulations (K) was set from 1 to 15 with 10 independent runs for each K. The burn-in length and Markov Chain Monte Carlo (MCMC) replication were set at 100,000 each run under the admixture and correlated allele frequency model. The STRUCTURE HARVESTER [\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e], a web-based program was used to summarize STRUCTURE output, visualize the likelihood values across multiple values of \u003cem\u003eK\u003c/em\u003e and, compute the natural logarithms of probability data \u003cem\u003e[LnP(K)\u003c/em\u003e] and the \u003cem\u003ead hoc\u003c/em\u003e statistic \u003cem\u003eΔK\u003c/em\u003e based on \u003cem\u003eEvanno\u003c/em\u003e method [\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe principal component analysis (PCA) was conducted using the Genomic Association and Prediction Integrated Tool (GAPIT) version 3[\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e] to detect the subpopulation structure present in the panel. The kinship matrix was estimated using the VanRaden algorithm [\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e] to measure the genetic relatedness among individuals in the association panel. The genetic relationship among the lines was determined based on the neighbor joining tree algorithm using the phylogenetic tree analysis in TASSEL software v5.2.93 [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. To determine the extent of linkage disequilibrium (LD), squared allele frequency correlations (\u003cem\u003er\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e) between all pairs of SNP markers were estimated using TASSEL software version 5.2.93 [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e]. To calculate the LD decay rate, the nonlinear regression model developed by [\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e], with modifications by Remington et al [\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e], was used to fit the LD decay curve into the scatterplot using the LOESS function in R.\u003c/p\u003e\n\u003ch3\u003eGenome-Wide Association Study\u003c/h3\u003e\n\u003cp\u003eGWAS analysis was carried out with 215,542 high quality SNPs (G) from 236 lines with BLUP values of eight phenotypic traits (GY, AD, SD, ASI, EPO, EPP, PH, and EH) using different multi-locus (ML) GWAS models under DS and optimum conditions. The first four PCAs and kinship (K) matrix were incorporated in the GWAS models as covariates to reduce false positives. The ML-GWAS was conducted with seven models including: (1) Fixed and random model circulating probability unification (FarmCPU)[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e], (2) mrMLM[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e], (3) FASTmrMLM [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e], (4) FASTmrEMMA [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e], (5) pLARmEB[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e], (6) pKWmEB[\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e], and (7) ISIS EM-BLASSO[\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. One multi-locus model was implemented in (GAPIT) R package software[\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e] and the other six multi-locus models were implemented in the \u003cem\u003emrMLM\u003c/em\u003e R package[\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. The nomenclature for naming QTN was done using the letter \"q\" to indicate QTN, followed by an abbreviation representing the trait name, underscoring the management conditions, the corresponding chromosome number, and the number of QTNs identified on that specific chromosome.\u003c/p\u003e \u003cp\u003eTo determine the genome-wide significant \u003cem\u003eP\u003c/em\u003e values threshold for the FarmCPU GWAS results, the effective number of independent SNPs (N) were calculated using the SimpM R program[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e, \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e] available on GitHub (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/LTibbs/SimpleM\u003c/span\u003e\u003cspan address=\"https://github.com/LTibbs/SimpleM\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The genome-wide significant \u003cem\u003eP\u003c/em\u003e value threshold was adjusted based on Bonferroni correction as the ratio of alpha value (α\u0026thinsp;=\u0026thinsp;0.05) divided by the effective number of independent SNPs (N\u0026thinsp;=\u0026thinsp;79,455). Hence, the genome-wide significant and suggestive levels were set as \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.05/N\u0026thinsp;=\u0026thinsp;6.29 \u0026times; 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e and \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1/N\u0026thinsp;=\u0026thinsp;1.26 \u0026times; 10\u0026thinsp;\u0026minus;\u0026thinsp;5, respectively, where \u003cem\u003eN\u003c/em\u003e is the effective number of independent SNPs. For multi-locus GWAS analysis, the genome-wide significant threshold was defined based on the threshold of LOD\u0026thinsp;\u0026ge;\u0026thinsp;3 (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0002). Manhattan plots and quantile-quantile plots were developed to visualize GWAS results using \u003cem\u003eCMplot R\u003c/em\u003e package (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/YinLiLin/CMplot\u003c/span\u003e\u003cspan address=\"https://github.com/YinLiLin/CMplot\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e).\u003c/span\u003e\u003c/p\u003e\n\u003ch3\u003eCandidate gene annotation and haplotype block analysis\u003c/h3\u003e\n\u003cp\u003eThe LD decay with a physical distance of 4.75 kb found in this study was used to find candidate genes. All the candidate genes for GY, AD, SD, PH, EH, EPO, and EPP located within regions from 4.75 kb upstream to 4.75 kb downstream associated with significant QTNs were identified and annotated using the B73 maize reference genome (B73 RefGen_V2)[\u003cspan additionalcitationids=\"CR79\" citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e]. The candidate gene annotations information was retrieved from the maizeGDB database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.maizegdb.org\u003c/span\u003e\u003cspan address=\"http://www.maizegdb.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe significant QTN, \u003cem\u003eqGY_DS1.1\u003c/em\u003e (S1_216149215) on chromosome 1 for GY located within the genomic regions of a candidate gene, \u003cem\u003eZm00001eb041070\u003c/em\u003e were extracted from the variant call format (VCF) using the site filtering option of VCF tools [\u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e]. The haplotype block analysis was then implemented in Haploview software version 4.2 [\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e] and \u003cem\u003egeneHapR\u003c/em\u003e [\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e]. The blocks were defined according to the criteria described by Gabriel et al. [\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e]. One-way Analysis of variance (ANOVA), boxplot, and multiple comparisons of phenotypic differences among haplotypes were implemented in \u003cem\u003eagricolae R package\u003c/em\u003e using Tukey\u0026rsquo;s Honestly Significant differences (HSD) test.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eGenomic-wide prediction\u003c/h2\u003e \u003cp\u003eGP was carried out on the 236 lines based BLUE values for traits across environments within management using ridge regression best linear unbiased prediction (\u003cem\u003errBLUP\u003c/em\u003e) model available in the \u003cem\u003errBLUP R package\u003c/em\u003e [\u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e]. Genomic estimated breeding values (GEBVs) were estimated using a five-fold cross-validation scheme by randomly sampling 80% and 20% of maize lines as training and testing sets, respectively. The prediction accuracy of the model was computed as the average Pearson\u0026rsquo;s correlation coefficient (r) between GEBV estimates from the training and testing set with 100 iterations. Bar plot was generated for each trait to visualize the means and standard deviation of prediction accuracy using the \u003cem\u003eggplot2 R package\u003c/em\u003e [\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003ePhenotypic variation, Descriptive statistics, and Correlation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe combined analysis of variance revealed significant (\u003cem\u003ep\u003c/em\u003e\u0026lt;0.05) genotype and GEI variations for GY, AD, SD, ASI, EPO, EPP, EH, and PH under DS and optimum conditions (Table 1). The mean, minimum, and maximum values of GY, AD, SD, ASI, EPO, EPP, EH, and PH revealed large variability for each trait under both DS and optimum conditions (Table 1). GY under DS was reduced by 71%. PH (214.69 cm) and EH (116.88 cm) were reduced significantly under DS compared to the\u0026nbsp;mean of PH (225.94 cm) and EH (112.27 cm) under optimum conditions. DS had a relatively small effect on EPO and EPP. DS conditions increased the average number of days for AD and SD compared to optimum conditions. Consequently, prolonging the interval between AD and SD (known as ASI) under drought by an average of 3 days compared to optimum conditions. The broad-sense heritability ranged from 0.25 for GY to 0.87 for EPO under DS and varied from 0.21 for EPP to 0.74 for AD under optimum conditions. The frequency distribution of each of the traits under the two experimental conditions (DS and optimum) is shown in Figure 1 where most of the traits showed continuous distributions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003ePearson’s correlation analysis showing the relationships among traits under DS and optimum conditions is presented in Figure 2. The correlation coefficients among the eight traits ranged from -0.29 to 0.93 under optimum conditions, whereas under DS conditions ranged from -0.56 to 0.86. Under optimum conditions, the highest significant positive coefficients correlation (r=0.93**) was observed between flowering traits, SD_OPT and AD_OPT, followed by PH_OPT and EH_OPT (r=0.78), GY_OPT and PH_OPT (r=0.65). GY_OPT had a significant positive correlation with PH_OPT, EH_OPT, and EPP_OPT while a significant but negative correlation was observed between GY_OPT with SD_OPT and ASI_OPT. Similarly, ASI_OPT had a significant negative correlation with AD_OPT, PH_OPT, EH_OPT, and EPP_OPT. Under DS conditions, the highest positive correlations were 0.83 (between GY_DS and EPP_DS), 0.66 (SD_DS and AD_DS), 0.64 (EPO_DS and EH_DS), 0.63 (PH_DS and EH_DS), and 0.57 (EPO_DS and AD_DS). GY_DS was negatively and significantly correlated with AD_DS, SD_DS, ASI_DS, and EPO_DS, while positively and significantly correlated with PH_DS and EPP_DS. It was observed that ASI_DS was positively correlated with AD_DS, SD_DS, and EPO_DS and negatively correlated with PH_DS, EH_DS, and EPP_DS. Additionally, AD_DS was positively correlated with SD_DS, EH_DS and EPO_DS, but negatively correlated with PH_DS and EPP_DS. Overall, a strong correlation was observed between the AD and SD under both optimum and DS conditions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMarker distribution, Population structure, Phylogenetic tree, and Kinship\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe distribution of 215,542 SNPs across the genome is presented in Figure 3A and supplementary Table S2. The number of SNP markers on each chromosome ranged from 14,629 to 33, 874 SNPs. Chromosome 1 had the highest number of SNPs of 33, 874, whereas chromosome 10 had the lowest number of SNPs with 14,629. The density of SNPs per mega-base pair (Mbp) varied from 84.89 Mbp for chromosome 4 to 116.79 Mbp for chromosome 5 with a mean of 104.28 Mbp.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe population structure of 236 diverse maize lines was determined by Bayesian based model in STRUCTURE and PCA (Supplementary Figure S1a and Figure 3B). The optimum number of K was obtained by plotting the number of clusters (K) against delta K (Supplementary Figure S1a). The Bayesian structure analysis revealed the presence of three distinct subgroups within the 236 maize lines when K = 3 and ten distinct subgroups K = 10 (Supplementary Figure S1b). The genetic structure was further examined by PCA (Figure 3B). The results revealed the presence of four distinct subpopulations separated by PC1 (12.78 %), PC2 (9.11%), and PC3 (7.19 %). The first three PCs explained over 29 % of the total genetic variation among the diverse maize lines. A scree plot of variance explained against the corresponding PCs as shown in Figure 3C was used to determine the optimal number of PCs to retain. The scree plot revealed that an optimal number (K) of four PCs could be retained for GWAS. The phylogenetic tree based on the neighbor-joining method as shown in Figure 3d revealed that the 236 diverse maize lines can be clustered into four main groups (I=66, II=48, III=100, and IV=22) differentiated by the different colors (Supplementary Table S1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe kinship matrix is utilized to assess the relatedness among individuals by considering the extent of allele sharing. The pattern of red shading in the center of the kinship matrix (Figure 3F reflected the level of genetic relatedness among individuals, indicating the presence of four stratified population structures. In general, the results from the PCA, phylogenetic tree, and kinship matrix, show that the panel of 236 diverse maize lines can be divided into four subpopulations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLinkage Disequilibrium and GWAS analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe nonrandom association of alleles between two genetic loci was examined to provide valuable information in locating genes tightly linked to SNP markers associated with traits of interest. A rapid LD decay pattern was observed across the ten chromosomes (Figure 3F). At \u003cem\u003er\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 0.2, the LD decay distance between SNPs on the 10 chromosomes varied from 2.83 kb (Chr3) to 11.83 kb (Chr4) with an average genome-wide LD decay with physical distance between SNPs of 4.74 kb.\u003c/p\u003e\n\u003cp\u003eBased on the seven multi-locus GWAS models (FarmCPU, mrMLM, FastmrMLM, FASTmrEMMA, pLARmEB, pKWmEB and ISIS EM-BLASSO), 219 and 265 QTNs distributed across the 10 chromosomes were detected under DS and optimum conditions, respectively (Figure 4, Supplementary Table S3 and S4). The QQ-plots depicting the observed vs expected LOD [−log10 (\u003cem\u003ep\u003c/em\u003e), \u003cem\u003ep\u003c/em\u003e= 6.29e-07] distributions showed a significant deviation from the expected distribution (Supplementary Figure S2 and S3), indicating the presence of QTNs linked to GY, AD, SD, ASI, PH, EH, EPP and EPO, respectively under DS and optimum conditions. The Manhattan plot showed that several QTNs with Bonferroni-adjusted threshold of greater than 6.20 [−log10 (\u003cem\u003ep\u003c/em\u003e), \u003cem\u003ep\u003c/em\u003e= 6.29e-07] and LOD threshold of ≥ 3 (p = 0.0002), were associated with GY, AD, SD, ASI, PH, EH, EPP and EPO, respectively under DS and optimum conditions.\u003c/p\u003e\n\u003cp\u003eThe number of significant QTNs detected by multiple GWAS models varied across traits. Of these, the power of detection among the seven models followed FASTmrMLM (62 QTNs) \u0026gt; pLARmEB (45 QTNs) \u0026gt; ISIS EM-BLASSO (37 QTNs) \u0026gt; pKWmEB (36 QTNs) \u0026gt; FarmCPU (16 QTNs) \u0026gt; FASTmrEMMA (12 QTNs) \u0026gt; mrMLM (7 QTNs) under DS and \u0026nbsp; FASTmrMLM (68 QTNs) \u0026gt; pLARmEB (62 QTNs) \u0026gt; pKWmEB (42 QTNs) \u0026gt; FASTmrEMMA (29 QTNs) \u0026gt; FarmCPU (25 QTNs) \u0026gt;ISIS EM-BLASSO (24 QTNs) \u0026gt; mrMLM (15 QTNs) under optimum conditions (Fig. 4C and 4D). Of these QTNs, 46, 33, 31, 30, 29, 20, 17, and 13 were found to be associated with EPO, ASI, AD, GY, SD, EPP, EH and PH under DS, respectively. Under optimum conditions, 40, 37, 36, 36, 32, 25, and 24 QTNs were associated with EPO, EH, ASI, GY, EPP, PH, SD and AD, respectively (Fig. 4A and 4B). With the proportion of \u003cem\u003eR\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e (%) by individual QTNs accounting for 0.00 ~ 8.18, 0.60 ~ 16.06, 3.93 ~ 15.29, 1.43 ~30.68, 0.00 ~ 11.42, 0.75 ~ 10.08, 1.32 ~ 16.06 and 2.00 ~ 12.66 (%) of total phenotypic variation for EPO, EH, ASI, GY, EPP, PH, SD and AD, respectively (Supplementary Table S4, S5 and S6).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSignificant QTNs detected by at least two ML-GWAS models were considered reliable and stable. A sum of 172 QTNs comprising 77 QTNs under DS and 95 QTNs under optimum conditions were discovered to be significantly associated with EPO, ASI, AD, GY, SD, EPP, EH, and PH, respectively (Fig. 4E and 4F; Supplementary Table S6). Of these significant QTNs under DS, 9, 9, 9, 11, 6, 6, 7, and 20 reliable QTNs were detected for GY, AD, SD, ASI, PH, EH, EPP, and EPO, respectively. The highest \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e (%) value was observed for GY QTN \u003cem\u003eqGY_DS2.1\u003c/em\u003e [S2_194907656:\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e =5.07 %-11.07%], for AD\u003cem\u003e\u0026nbsp;qAD_DS1.2\u003c/em\u003e [S1_32228028: \u003cem\u003eR\u003csup\u003e2\u0026nbsp;\u003c/sup\u003e\u003c/em\u003e=5.27%-11.27%], for SD\u003cem\u003e\u0026nbsp;qSD_DS10.1\u003c/em\u003e [S10_2032146:\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 5.36 % - 9.92 %], for ASI\u003cem\u003e\u0026nbsp;qASI_DS5.1\u003c/em\u003e [S5_151886720:\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e =5.94 %-11.19 %], for PH\u003cem\u003e\u0026nbsp;qPH_DS3.2\u003c/em\u003e [S3_149683959:\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 9.62%-11.27 %], for EH\u003cem\u003e\u0026nbsp;qEH_DS2.1\u003c/em\u003e [S2_60655022:\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 6.23%-6.78%], for EPP \u003cem\u003eqEPP_DS1.1\u003c/em\u003e [S1_217115118:\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 9.19 %-13.26 %] and for EPO\u003cem\u003e\u0026nbsp;qEPH_DS2.1\u003c/em\u003e [S2_36458273:\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = \u0026nbsp;0.96 % - 6.15 %].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn addition, of the 95 QTNs detected in optimum condition, 12, 9, 6, 13, 13, 11, 18, and 13 QTNs were found to be associated with GY, AD, SD, ASI, EH, PH, EPO, and EPP, respectively. The QTN, \u003cem\u003eqGY_OPT5.2\u003c/em\u003e [S5_193538664: \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 5.34% - 10.19%] recorded the highest \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e (%) for GY. One QTN, \u003cem\u003eqAD_OPT8.1\u003c/em\u003e (S8_105322358) was shared by AD (\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 6.03 % -12.66 %) and SD (\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 3.99 %-16.06 %). For ASI, \u003cem\u003eqASI_OPT6.1\u003c/em\u003e [S6_159006932: \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 6.30% - 14.10%] explained the highest \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e (%). While \u003cem\u003eqEH_OPT10.4\u003c/em\u003e [S10_88760138: \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 9.39 % - 16.06 %] showed the highest \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e (%) for EH, \u003cem\u003eqPH_OPT3.1\u003c/em\u003e [S3_172201680: \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 1.49 % - 10.08%] for PH, \u003cem\u003eqEPO_OPT2.2\u003c/em\u003e [S2_233748945: \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 1.56 % - 3.26 %] for EPO and \u003cem\u003eqEPP_OPT8.2\u003c/em\u003e [S8_75600223: \u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/em\u003e = 0.002 % -5.48 %] for EPP (Figure 4 and Figure 5).\u003c/p\u003e\n\u003cp\u003eThe detected QTNs contributing revealed positive and negative allelic effects for different traits (Figure 4E and Figure 5). Among the 172 QTNs, few QTNs were detected under both DS and optimum conditions. For instance, two QTNs \u003cem\u003eqGY_DS2.1\u003c/em\u003e and \u003cem\u003eqGY_DS1.1\u003c/em\u003e explained the highest and positive effects with 0.34 and 0.24, respectively under DS for GY. On the other hand, two new QTNs \u003cem\u003eqGY_OPT4.1\u003c/em\u003e and \u003cem\u003eqGY_OPT10.1\u003c/em\u003e showed the highest and positive effects of 0.56 and 0.49, respectively for GY under optimum conditions. Negative QTN effects are desirable for flowering traits (AD, SD, and ASI) under DS for selecting earliness. For instance, QTNs \u003cem\u003eqAD_DS2.1\u003c/em\u003e, \u003cem\u003eqSD_DS8.1\u003c/em\u003e, \u003cem\u003eqASI_DS1.1\u003c/em\u003e exhibited large and negative effects for AD (-1.72), SD (-1.79), and ASI (-0.77) on chromosome 2, 8 and 1, respectively (Figure 4E). The distribution of QTNs (Figure 5) on the 10 chromosomes showed that chromosome 5 captured the higher number of reliable QTNs (12 under DS and 14 under optimum conditions). Two QTNs, \u003cem\u003eqPH_OPT2.2\u003c/em\u003e and \u003cem\u003eqEH_OPT2.1\u003c/em\u003e exhibited a pleiotropic effect on both PH and EH under optimum conditions.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCandidate gene identification and Annotation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCandidate genes analysis was conducted for significant QTNs identified in this study to elucidate the molecular, biological, and physiological mechanisms controlling traits under DS and optimum conditions. A total of 43 candidate genes were discovered and annotated, among them 18 and 25 candidate genes were\u0026nbsp;identified under DS and optimum conditions, respectively (Table 2). Two candidate genes closely associated with the QTNs for improved GY were identified under DS, namely, \u003cem\u003eZm00001eb041070\u003c/em\u003e and \u003cem\u003eZm00001d045665\u003c/em\u003e (Table 2). Nine candidate genes potentially associated with flowering traits (AD, SD and ASI) were identified under DS (Table 2). Of these five candidate genes were associated with ASI under DS. Similarly, four and three candidate genes were found to be associated with EPO and EPP, respectively. Under optimum conditions, 3, 5, 8, 3, and 3 candidate genes were identified for GY, PH, EH, EPO, and EPP, respectively (Table 2). On the other hand, one candidate gene each was identified for AD, SD, and ASI.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHaplotype Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe QTN \u003cem\u003eqGY_DS1.1\u003c/em\u003e (S1_216149215) associated with GY under DS was identified in a 216.15 Mb region on chromosome 1 based on pairwise LD correlation (Figure 6A and 6C). The QTN region contains six distinct SNPs with relatively high LD (Figure 6C). Four major haplotypes were detected among the 236 lines with Haplotypes frequencies of 142 (60 %), 39 (17 %), 29 (12 %), and 7 (3%) for Hap1 (GAGGGC), Hap2 (AAGGGC), Hap3 (GTAATG), and Hap4 (GAGGTG), respectively (Figure 6B). A significant difference was observed between haplotypes Hap1 and Hap2 (Figure 6D). The mean GY was 2.2 t/ha for Hap1, 1.80 t/ha for Hap2, 2.10 t/ha for Hap3, and 2.2 t/ha for Hap4, respectively. (Figure 6D). Hap1 was considered a superior haplotype since it contributes to the highest mean performance compared to other haplotypes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenomic prediction\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAmong the GP models, RR-BLUP is computationally less intensive and is well suited for routine application in plant breeding trials. Therefore, we used the RR-BLUP model to estimate the performance of maize genotypes for various traits under optimum and DS conditions (Figure 7). Prediction accuracies were moderate to high for all eight traits under optimum and low to moderate under DS conditions (Figure 7). The observed prediction accuracy for GY, AD, SD, ASI, PH, EH, EPO, and EPP were 0.29, 0.58, 0.45, 0.44, 0.61, 0.54, 0.24, and 0.25, respectively under DS conditions, and 0.65, 0.72, 0.69, 0.37, 0.71, 0.51, 0.48, and 0.41, under optimum conditions.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eDrought has long been recognized as a major abiotic factor limiting crop growth and productivity [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e]. Genetic dissection of the genomic regions responsible for GY and other secondary traits under DS will allow breeders to improve their breeding efficiency in the development of climate-resilient varieties and, also, facilitate the introgression of the favorable alleles into elite germplasm using marker-assisted selection [\u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e]. Furthermore, understanding the complex genetic basis of drought tolerance, GY and other secondary traits facilitate the opportunity to test for the indirect selection of GY under DS [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn the present study, significant genotypic variance and wide range phenotypic variation observed for all traits under optimal and DS, indicated the presence of adequate genetic diversity within the GWAS panel and that progress from selection for GY and secondary traits could be achieved through breeding. These findings support earlier studies [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e, \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e], which indicated the existence of sufficient genetic variability. The notable significant GEI in the study implied that the testing sites, drought and optimal growing conditions were discriminating enough in identifying genotypic differences in the response of the GWAS panel to drought and optimal conditions, and that these differences can be largely attributed to the differences in environmental factors, such as soil types, temperature, and amount of rainfall [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Heritability estimates are crucial for ascertaining the effectiveness and progress anticipated from future phenotypic selection for yield and secondary traits under DS. The lower heritability estimates for GY under drought stress indicated that the correlated secondary trait, ASI with high heritability and strong correlation with GY could enhance the effectiveness of the selection response [\u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e92\u003c/span\u003e]. These findings are consistent with earlier studies [\u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e, \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e, \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e, \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e94\u003c/span\u003e]. A study by Ndlovu et al. [\u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e95\u003c/span\u003e] on multiple bi-parental maize populations in Kenya under water-stressed and well-watered conditions also reported lower heritability for GY and low genotypic variance under DS conditions. The notably significant and negative correlation between GY and ASI under DS indicated that ASI is a suitable secondary trait to facilitate the selection of GY and a target for improving drought tolerance in maize. These findings are consistent with earlier studies [\u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e94\u003c/span\u003e, \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e96\u003c/span\u003e, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e].\u003c/p\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eLinkage disequilibrium, population structure and association mapping\u003c/h2\u003e \u003cp\u003eThe LD decay distance determines the power of the GWAS [\u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e]. The high-resolution provided by GWAS is largely influenced by the nature of LD and the extent of its decay across the genome [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e], which is population specific and influenced by recombination rate, number of generations of recombination, genetic drift, selection within populations, and population admixture [\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e, \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e]. In this study, genome-wide LD analysis revealed that the GWAS panel decayed rapidly at 4.75 kb (r\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.2). This suggests that the GWAS panel contained adequate genetic diversity suitable for GWAS analysis resulting from the historical and evolutionary recombination events present in the panel. Previous studies conducted by Rashid et al .[\u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e100\u003c/span\u003e], reported a rapid LD decay of 0.9 kb at r\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.2 for CIMMYT Asia Association Mapping, 1.75 kb at r\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.2 for Drought Tolerant Maize for Africa (DTMA) and 0.99 kb at r\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.2 for Improved Maize for African Soils (IMAS) panels. Lu et al. [\u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e101\u003c/span\u003e], reported that LD decay distance in temperate germplasm (10 to 100 kb), was two to ten times greater than LD decay distance in tropical germplasm (5 to 10 kb), suggesting that tropical maize has significant genetic diversity for breeding programs. The population structure within the GWAS panel was best explained when separated into four groups. The incorporation of the four PCs in GWAS analysis was sufficient to reduce false positive marker-trait association in the present study as evident in the Q-Q plots and Manhattan plots.\u003c/p\u003e \u003cp\u003eThe variation in the number of QTNs detected across the GWAS models suggests that the differences can be attributed to genetic algorithms implemented in the different models. A similar conclusion that multi-locus models show superior performance over single-locus models in terms of statistical power in detecting QTNs, particularly in the accuracy of QTN effect estimation and reducing false positive rates has been found in many other studies [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e102\u003c/span\u003e, \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e]. This further implied that adopting multiple GWAS models could help to complement each other in identifying reliable and stable QTNs [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. The proportion of phenotypic variance explained R\u003csup\u003e2\u003c/sup\u003e (%) by each QTNs and the QTN effects observed for GY, EPP and EPO in this study were relatively low under DS. The results implied that the GY, EPP and EPO under DS are likely influenced by numerous QTNs with small effect. Similar results were reported in earlier study [\u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eCandidate genes discovery\u003c/h2\u003e \u003cp\u003eIn the present study, two candidate genes (\u003cem\u003eZm00001eb041070 and Zm00001d045665\u003c/em\u003e) significantly associated with GY were identified under DS. The peak QTNs, \u003cem\u003eqGY_DS1.1\u003c/em\u003e is located within the gene, \u003cem\u003eZm00001eb041070\u003c/em\u003e encoding AP2 (APETALA2)-EREBP (Ethylene Responsive Element Binding Protein) transcription factor 60 on chromosome 1. It plays a critical role in regulating genes involved in plant growth and development, biotic, and abiotic stress [\u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e105\u003c/span\u003e]. Ningning et al [\u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e106\u003c/span\u003e] reported three candidate genes, \u003cem\u003eGRMZM2G322672 (EREB37), GRMZM2G026926 (ERF), and GRMZM2G169654 (RAV)\u003c/em\u003e, belonging to the AP2/EREBP family that plays important roles in maize response to DS. In maize, \u003cem\u003eZmEREBP60\u003c/em\u003e is a positive regulator under DS and its expression, significantly upregulated in the roots, coleoptiles, and leaves in response to drought [\u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e107\u003c/span\u003e, \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e108\u003c/span\u003e] Another gene \u003cem\u003eZm00001eb041070\u003c/em\u003e reported to control AD under DS [\u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e, \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e109\u003c/span\u003e]. In addition, a gene \u003cem\u003eZm00001d045665\u003c/em\u003e encoding Ubiquitin receptor Radiation sensitivity proteins-23 (RAD23) was identified \u0026minus;\u0026thinsp;46 bp around the peak QTN, \u003cem\u003eqGY_DS.9.1\u003c/em\u003e. This receptor is known to participate in DNA repair and play a major role in the cell cycle, morphology, and fertility of plants through their delivery of UPS substrates to the 26S proteasome [\u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e110\u003c/span\u003e, \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e111\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe QTNs associated with GY, \u003cem\u003eqGY_OPT2.1\u003c/em\u003e, \u003cem\u003eqGY_OPT1.1\u003c/em\u003e, \u003cem\u003eqGY_OPT10.1\u003c/em\u003e are nearest to three candidate genes, \u003cem\u003eZm00001d005785, Zm00001eb034820 and Zm00001eb425230\u003c/em\u003e which encode pentatricopeptide (PPR) repeat-containing protein [\u003cspan citationid=\"CR112\" class=\"CitationRef\"\u003e112\u003c/span\u003e, \u003cspan citationid=\"CR113\" class=\"CitationRef\"\u003e113\u003c/span\u003e], PfkB-like carbohydrate kinase family protein[\u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e114\u003c/span\u003e] and Trihelix-transcription factor (TTF) 25 [\u003cspan citationid=\"CR115\" class=\"CitationRef\"\u003e115\u003c/span\u003e, \u003cspan citationid=\"CR116\" class=\"CitationRef\"\u003e116\u003c/span\u003e], respectively. In a QTN-by environment interaction GWAS under optimum, DS, heat stress, and combined drought and heat stress, Wen et al [\u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e] identified \u003cem\u003eGRMZM2G110851\u003c/em\u003e related to GY around the locus S1_299093763 encoding pentatricopeptide repeat protein, required for mitochondrial function and kernel development in maize. In rice and Arabidopsis, phosphofructokinase B-type carbohydrate kinase family protein, PFKB1 regulates leaf and plant growth by controlling chloroplast biogenesis[\u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e114\u003c/span\u003e, \u003cspan citationid=\"CR117\" class=\"CitationRef\"\u003e117\u003c/span\u003e]. The TTFs are light-responsive proteins and play a critical role in plant growth and development and stress tolerance in maize [\u003cspan citationid=\"CR116\" class=\"CitationRef\"\u003e116\u003c/span\u003e, \u003cspan citationid=\"CR118\" class=\"CitationRef\"\u003e118\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eFlowering traits in maize are quantitative in nature and regulated by complex genetic mechanisms [\u003cspan citationid=\"CR119\" class=\"CitationRef\"\u003e119\u003c/span\u003e]. Several candidate genes associated with flowering traits under DS and optimal conditions have been reported in maize[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e, \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e, \u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e120\u003c/span\u003e, \u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e121\u003c/span\u003e]. Here, some new candidates have been found to be associated with flowering traits. Rubisco large subunit methyltransferase encoded by \u003cem\u003eZm00001eb058200\u003c/em\u003e on chromosome 1 near peak QTN \u003cem\u003eqAD_DS1.1\u003c/em\u003e was found under DS. It is a key photosynthetic enzyme in the chloroplast that catalyzes the fixation of atmospheric CO\u003csub\u003e2\u003c/sub\u003e during photosynthesis in the carbon assimilation pathway and drought-related responses [\u003cspan citationid=\"CR122\" class=\"CitationRef\"\u003e122\u003c/span\u003e, \u003cspan citationid=\"CR123\" class=\"CitationRef\"\u003e123\u003c/span\u003e]. The gene \u003cem\u003eZm00001eb010210\u003c/em\u003e, associated with AD under DS encodes plant U-box type E3 ubiquitin ligase [\u003cspan citationid=\"CR124\" class=\"CitationRef\"\u003e124\u003c/span\u003e, \u003cspan citationid=\"CR125\" class=\"CitationRef\"\u003e125\u003c/span\u003e]. The ligase gene family plays a key role in the abiotic stress adaptation in maize [\u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e126\u003c/span\u003e]. Another candidate gene for AD, \u003cem\u003eZm00001eb012590\u003c/em\u003e encodes Trifunctional UDP-glucose 46-dehydratase, an enzyme plays a key role in nucleotide sugar biosynthetic pathway, maintaining cell wall integrity during maize cell growth and response to abiotic stress[\u003cspan citationid=\"CR127\" class=\"CitationRef\"\u003e127\u003c/span\u003e, \u003cspan citationid=\"CR128\" class=\"CitationRef\"\u003e128\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eASI is a key index in breeding programs for facilitating indirect selection for GY under DS. Maize exhibits postponement in the silking process under stress, resulting in an extended period between AD and SD [\u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e]. Five candidate genes \u003cem\u003eZm00001eb047160, Zm00001eb364110, Zm00001eb430770, Zm00001eb426690 and Zm00001d023240\u003c/em\u003e associated with ASI under DS, encodes for kaurene synthase1[\u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e129\u003c/span\u003e], \u003cem\u003eTEOSINTE BRANCHED1/CYCLOIDEA/PROLIFERATING CELL FACTOR (TCP)\u003c/em\u003e -transcription factor 20 [\u003cspan citationid=\"CR130\" class=\"CitationRef\"\u003e130\u003c/span\u003e], Ribosomal protein s11-beta [\u003cspan citationid=\"CR131\" class=\"CitationRef\"\u003e131\u003c/span\u003e], Hexosyltransferase[\u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e132\u003c/span\u003e] and Serine/threonine-protein kinase BLUS1[\u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e133\u003c/span\u003e], respectively. Additionally, the candidate gene, \u003cem\u003eZm00001eb047160\u003c/em\u003e associated with PH under optimal conditions, encodes kaurene synthase, an enzyme responsible for the biosynthesis of gibberellins and phytoalexin metabolism [\u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e129\u003c/span\u003e, \u003cspan citationid=\"CR134\" class=\"CitationRef\"\u003e134\u003c/span\u003e]. The TCP family of transcription factors participates in the regulation of cell growth, proliferation, and response to abiotic stress [\u003cspan citationid=\"CR135\" class=\"CitationRef\"\u003e135\u003c/span\u003e]. In maize, \u003cem\u003eZmGA20ox3\u003c/em\u003e play a crucial in regulating PH and enhancing drought tolerance [\u003cspan citationid=\"CR136\" class=\"CitationRef\"\u003e136\u003c/span\u003e]. Furthermore, candidate genes, \u003cem\u003eZm00001eb364110\u003c/em\u003e and \u003cem\u003eZm00001eb408080\u003c/em\u003e related to ASI and EPP on chromosome 8 and 10, respectively, under drought and optimal conditions encode TCP-transcription factor [\u003cspan citationid=\"CR137\" class=\"CitationRef\"\u003e137\u003c/span\u003e, \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e138\u003c/span\u003e]. Previous studies reported that \u003cem\u003eZmTCP\u003c/em\u003e family proteins, particularly \u003cem\u003eZmTCP42\u003c/em\u003e, is a key positive regulator of drought tolerance in maize. However, overexpression of the \u003cem\u003eZmTCP14 gene\u003c/em\u003e significantly reduced the tolerance to drought [\u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e138\u003c/span\u003e]. However, \u003cem\u003eZmTCP14\u003c/em\u003e gene-edited plants showed improved drought tolerance and GY[\u003cspan citationid=\"CR137\" class=\"CitationRef\"\u003e137\u003c/span\u003e, \u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e138\u003c/span\u003e]. Additionally, the candidate gene \u003cem\u003eZm00001d023240\u003c/em\u003e encodes serine/threonine-protein kinase, \u003cem\u003eBLUE LIGHT SIGNALLING 1 (BLUS1)\u003c/em\u003e, play a crucial role in blue light-induced stomatal opening and transduction of signals related to abiotic stress [\u003cspan citationid=\"CR137\" class=\"CitationRef\"\u003e137\u003c/span\u003e, \u003cspan citationid=\"CR139\" class=\"CitationRef\"\u003e139\u003c/span\u003e]. This suggests that serine/threonine-protein kinase BLUS1 participates in regulating ASI during DS.\u003c/p\u003e \u003cp\u003eAmong the candidate genes identified with association to EPO under DS, \u003cem\u003eZm00001d044508\u003c/em\u003e encodes cysteine-rich receptor kinase, which plays a critical role in plant immunity, and abiotic stress response [\u003cspan citationid=\"CR113\" class=\"CitationRef\"\u003e113\u003c/span\u003e, \u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e114\u003c/span\u003e]. Another gene \u003cem\u003eZm00001eb426230\u003c/em\u003e encodes GDSL-motif esterase/acyltransferase/lipase, play a key role in cuticle development, seed oil storage, and biotic and abiotic stress responses [\u003cspan additionalcitationids=\"CR143\" citationid=\"CR142\" class=\"CitationRef\"\u003e142\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR144\" class=\"CitationRef\"\u003e144\u003c/span\u003e]. The \u003cem\u003eZm00001d026371\u003c/em\u003e gene associated with EPO encodes anthocyanidin reductase, a key enzyme in involved in mediating proanthocyanidin and lignin biosynthesis in response to DS [\u003cspan citationid=\"CR145\" class=\"CitationRef\"\u003e145\u003c/span\u003e, \u003cspan citationid=\"CR146\" class=\"CitationRef\"\u003e146\u003c/span\u003e]. Previous study [\u003cspan citationid=\"CR147\" class=\"CitationRef\"\u003e147\u003c/span\u003e] reported that the anthocyanidin reductase enzyme system plays a key role in reducing excess reactive oxygen species (ROS) induced by DS, which can damage cellular membranes and cause cell death in maize. The \u003cem\u003eZm00001d002247\u003c/em\u003e gene, which underpinned an association with EPP under DS, encodes the MATH-BTB domain containing proteins, which play a key role in abiotic stress response [\u003cspan citationid=\"CR148\" class=\"CitationRef\"\u003e148\u003c/span\u003e, \u003cspan citationid=\"CR149\" class=\"CitationRef\"\u003e149\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe candidate gene \u003cem\u003eZm00001d001877\u003c/em\u003e near peak QTN, \u003cem\u003eqEH_OPT2.1\u003c/em\u003e and \u003cem\u003eqPH_OPT2.2\u003c/em\u003e exhibited pleiotropic effect on EH and PH under optimal conditions. This implied that a positive correlation between these two traits and that the two QTNs simultaneously regulated [\u003cspan citationid=\"CR150\" class=\"CitationRef\"\u003e150\u003c/span\u003e]. The candidate gene, \u003cem\u003eZm00001eb431080\u003c/em\u003e on chromosome 10 exhibited pleiotropic effect with PH and EPO under optimum conditions. This candidate gene encodes basic Helix-Loop-Helix (\u003cem\u003ebHLH\u003c/em\u003e) transcription factor, involved in anthocyanin biosynthesis and response to biotic and abiotic stresses [\u003cspan additionalcitationids=\"CR152\" citationid=\"CR151\" class=\"CitationRef\"\u003e151\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR153\" class=\"CitationRef\"\u003e153\u003c/span\u003e]. ABA-responsive protein encoded by \u003cem\u003eZm00001d023664\u003c/em\u003e is associated with EPP plays a critical role in controlling stomatal closure, and regulate plants response to abiotic stress [\u003cspan citationid=\"CR154\" class=\"CitationRef\"\u003e154\u003c/span\u003e, \u003cspan citationid=\"CR155\" class=\"CitationRef\"\u003e155\u003c/span\u003e]. The candidate gene, \u003cem\u003eZm00001eb003210\u003c/em\u003e associated with EPP encodes CONSTANS-LIKE (COL) TIMING OF CAB1 protein domain, which plays a crucial role in regulating flowering through photoperiodic control and response to abiotic stress [\u003cspan citationid=\"CR156\" class=\"CitationRef\"\u003e156\u003c/span\u003e, \u003cspan citationid=\"CR157\" class=\"CitationRef\"\u003e157\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eHaplotype estimation\u003c/h2\u003e \u003cp\u003eHaplotype refers to a set of alleles combinations within a specific genomic region showing significant LD and are inherited together [\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e, \u003cspan citationid=\"CR158\" class=\"CitationRef\"\u003e158\u003c/span\u003e]. Identification of superior haplotypes linked to functional genes is important for developing markers to accelerate genetic gain and facilitate efficient selection in breeding operations. The haplo-pheno analysis identified Hap1 (GAGGGC) as a superior haplotype with high mean performance associated GY near QTN, \u003cem\u003eqGY_DS1.1\u003c/em\u003e (\u003cem\u003eS1_216149215\u003c/em\u003e) in the present study. Incorporating this superior haplotype in maize breeding will facilitate the selection of improved drought tolerant maize cultivars. These findings revealed that the GY of lines carrying superior haplotypes can exhibit improved drought tolerance compared to lines with unfavorable haplotypes. This is evident in their ability to maintain higher yields in DS and, these lines can also be utilized as parents for hybridization and backcrossing programs in the future.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eGenome-wide prediction\u003c/h2\u003e \u003cp\u003eGS has been successfully applied in maize for complex traits like GY under optimum, low soil N stress, heat stress and DS conditions [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e, \u003cspan citationid=\"CR159\" class=\"CitationRef\"\u003e159\u003c/span\u003e]. In the present study, we compared the prediction accuracies under optimum and DS conditions (Fig.\u0026nbsp;7). The prediction accuracies were moderate to high for GY and other traits under optimum conditions whereas the accuracies were slightly lower under DS conditions. The accuracy observed for all traits under both optimum and DS conditions reveal the effect of heritability as the traits with higher heritability generally had higher prediction accuracy. The observed prediction accuracies for GY and other traits are comparable to earlier studies reported under different stresses in maize [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e, \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e95\u003c/span\u003e, \u003cspan citationid=\"CR159\" class=\"CitationRef\"\u003e159\u003c/span\u003e, \u003cspan citationid=\"CR160\" class=\"CitationRef\"\u003e160\u003c/span\u003e]. In genome-wide predictions, the less complex traits like AD, SD and PH had higher accuracy compared to the GY, which is consistent with the nature of trait complexity [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eBreeding for drought tolerance is complex and very expensive. The observed prediction accuracy for GY is 0.29 and 0.65 under DS and optimum conditions, respectively. On the other hand, the phenotypic selection efficiency is 0.50 (square root of heritability of GY) under DS and 0.73 under optimum conditions. Nevertheless, with the possibility to complete three cycles per year and requirement of lesser resources for implementing GS compared to phenotypic selection, endorse GS to integrate to improve the selection efficiency for drought tolerance as an attractive option in a longer run objective of breeding program. Further, integration of GS with GWAS results leads not only to an increase in the prediction accuracy, but also helps to validate the function of the identified candidate genes and increases in the accumulation of favorable alleles for drought tolerance in improved set of elite lines. Further GS can remarkably reduce the resources required for selection and improve breeding efficiency.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eTremendous variations for GY and secondary traits associated with DS existed in the maize GWAS panel used in this study. Based on seven GWAS models, a total of 172 stable and reliable QTNs (77 under DS and 95 under optimal conditions) were identified. Among these QTNs two QTNs, \u003cem\u003eqGY_DS1.1\u003c/em\u003e for GY and \u003cem\u003eqASI_DS8.2\u003c/em\u003e for ASI were novel. This result demonstrated that combining different GWAS models is effective and powerful in exploiting the complementary strengths of different methods in identifying QTNs with small and large effects for GY and secondary traits with a complex genetic basis and low heritability under DS. In addition, a total of 43 candidate genes were detected (18 under DS and 25 optimum conditions). Among these, two key candidate genes are closely associated with GY and nine closely associated with flowering traits under DS. Genomic prediction revealed moderate to high accuracies under optimum and DS conditions. Integration of GS with GWAS results leads not only to an increase in prediction accuracy, but also helps to increase in the accumulation of favorable alleles for drought tolerance. Overall stable QTNs and candidate genes for GY and secondary trait detected in the current study can serve as invaluable resources for improving GY under DS in maize. The identified superior haplotype can be converted to a breeder-friendly marker to increase its favorable alleles in elite breeding lines. These findings provide valuable insight into the genetic basis of drought tolerance for GY and secondary traits in tropical maize.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eANOVA:\u003c/em\u003e\u003c/strong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Analysis of variance\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eAD:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAnthesis days\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eASI:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eAnthesis-silking interval\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eBLINK:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eBayesian-information and linkage-disequilibrium iteratively nested keyway\u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eCV:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eCoefficient of variation\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eCIMMYT:\u003c/em\u003e\u003c/strong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;International Maize and Wheat Improvement Centre\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eDS:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eDrought Stress\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eECMLM:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eEnriched compressed mixed linear model\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eEMMAX:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eEfficient mixed model association eXpedited\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eEPO:\u003c/em\u003e\u003c/strong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Ear position\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eEH:\u003c/em\u003e\u003c/strong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Ear height\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eEPP:\u003c/em\u003e\u003c/strong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Ear per plant\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFarmCPU:\u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003eFixed and random model Circulating Probability Unification\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFaST-LMM:\u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003eFactored spectrally transformed linear mixed models\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFASTmrEMMA:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eFast multi-locus random-SNP-effect efficient mixed model analysis\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eFASTmrMLM:\u003c/em\u003e\u003c/strong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Fast mrMLM\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eGP\u003c/em\u003e\u003c/strong\u003e:\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Genomic prediction\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eGS:\u003c/em\u003e\u003c/strong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Genomic selection\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eGY:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eGrain yield\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eGLM:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eGeneral linear model\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eGWAS:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eGenome-wide association study\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eGEMMA:\u003c/em\u003e\u003c/strong\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Genome-wide efficient mixed-model association\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eISIS EM-BLASSO:\u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003eIterative modified-sure independence screening expectation-maximization-Bayesian LASSO\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eLD:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003eLinkage disequilibrium\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eLOD:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eLogarithm of Odds\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eML-GWAS:\u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eMulti-locus genome-wide association studies\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eMLM:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003eMixed linear model\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003emrMLM:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003emulti-locus random-SNP-effect MLM\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003ePCA:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003ePrincipal component analysis\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003epLARmEB:\u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e polygenic-background-control-based least angle regression plus empirical Bayes\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003epKWmEBb:\u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eIntegration of Kruskal-Wallis test with empirical Bayes\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eQTL:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003eQuantitative trait locus\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eQTN:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eQuantitative trait nucleotide\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eR\u003csup\u003e2\u003c/sup\u003e:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eThe proportion of total phenotypic variance explained by each QTN\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003er\u003csup\u003e2\u003c/sup\u003e:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eThe squared correlation coefficient\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eSL-GWAS:\u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eSingle-locus genome-wide association studies\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eSNP:\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/em\u003e\u003c/strong\u003e Single-nucleotide polymorphic\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData Availability\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data generated or analyzed during this study are included in this published article and its supplementary information files.\u003c/p\u003e\n\u003cp\u003eAll relevant data are within the manuscript and its supplementary files.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur sincere thanks to CIMMYT\u0026rsquo;s research technicians in Kenya. The authors are grateful to the International Maize and Wheat Improvement Center (CIMMYT) and partner scientists and technicians who generated the germplasm and designed and conducted the experiments that we used to explore our objectives.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe research was supported by the Bill and Melinda Gates Foundation (B\u0026amp;MGF), Foundation for Food and Agriculture Research (FFAR), and the United States Agency for International Development (USAID) through AG2MW (Accelerating Genetic Gains in Maize and Wheat for Improved Livelihoods, B\u0026amp;MGF Investment ID INV-003439), the Stress Tolerant Maize for Africa (STMA, B\u0026amp;MGF Grant # OPP1134248) project, and the CGIAR Research Program on Maize (MAIZE). This research was part of the PhD thesis of the first author, funded by the WACCI ACE Impact project at West Africa Centre for Crop Improvement (WACCI), University of Ghana, Legon, Ghana.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors and Affiliations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInternational Maize and Wheat Improvement Center (CIMMYT), C/O: World Agroforestry Centre (ICRAF), United Nations Avenue, Gigiri, P.O. Box 1041-00621, Nairobi, Kenya.\u003c/p\u003e\n\u003cp\u003eManigben Kulai Amadu, Yoseph Beyene, Vijay Chaikam, Boddupalli M Prasanna and Manje Gowda\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e2\u003c/sup\u003eWest Africa Centre for Crop Improvement (WACCI), University of Ghana, PMB 30 Legon, Accra, Ghana\u003c/p\u003e\n\u003cp\u003eManigben Kulai Amadu, Pangirayi B. Tongoona, Eric Y. Danquah and Beatrice E. Ifie\u003c/p\u003e\n\u003cp\u003eCSIR-Savanna Agricultural Research Institute, PO. Box 52, Tamale, Nyankpala, Ghana\u003c/p\u003e\n\u003cp\u003eManigben Kulai Amadu,\u003c/p\u003e\n\u003cp\u003eInternational Maize and Wheat Improvement Center (CIMMYT), Km 45, Carretera M\u0026eacute;xico-Veracruz, CP 52640, Edo. de M\u0026eacute;xico, Mexico\u003c/p\u003e\n\u003cp\u003eJuan Burgueno\u003c/p\u003e\n\u003cp\u003eInstitute of Biological, Environmental \u0026amp; Rural Sciences (IBERS), Aberystwyth University, Aberystwyth, SY23 3EE, Wales, UK\u003c/p\u003e\n\u003cp\u003eBeatrice E. Ifie\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eContributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization, M.K.A., M.G., Y.B. and P.B.T.; methodology, M.G.,M.K.A., Y.B., V.C., and P.B.T; software, M.G., J.B, and M.K.A.; validation, M.K.A., M.G., Y.B., V.C., and P.B.T.; formal analysis, M.K.A.,M.G., J.B; investigation, Y.B., M.G. and M.K.A.; resources, B.M.P., M.G., E.Y.B. and V.C.; data curation, M.K.A., M.G. and Y.B.; writing\u0026mdash;original draft preparation, M.K.A. and M.G.; writing\u0026mdash;review and editing, M.K.A., M.G.,V.C, Y.B., B.M.P., J.B, P.B.T., and B.E.I., visualization, M.G. and Y.B.; supervision, M.G., P.B.T., E.Y.D., B.E.I. and B.M.P.; project administration, M.G. and Y.B.; funding acquisition, B.M.P and E.Y.D. \u0026nbsp;All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclarations Ethics approval and consent to participate\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll experimental studies on plants complied with relevant institutional, national, and international guidelines and legislation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eErenstein O, Jaleta M, Sonder K, Mottaleb K, Prasanna BM. Global maize production, consumption and trade: trends and R\u0026amp;D implications. Food Secur. 2022;:1295\u0026ndash;319.\u003c/li\u003e\n\u003cli\u003eThe Global Action for Fall Armyworm Control: Action framework 2020\u0026ndash;2022. 2020.\u003c/li\u003e\n\u003cli\u003ePrasanna BM, Nair SK, Babu R, Gowda M, Zhang X, Xu Y, et al. Increasing genetic gains in maize in stress-prone environments of the tropics. In: Genomic Designing of Climate-Smart Cereal Crops. 2020.\u003c/li\u003e\n\u003cli\u003eTarekegne A, Wegary D, Cairns JE, Zaman-Allah M, Beyene Y, Negera D, et al. Genetic gains in early maturing maize hybrids developed by the International Maize and Wheat Improvement Center in Southern Africa during 2000\u0026ndash;2018. Front Plant Sci. 2023;14:1321308.\u003c/li\u003e\n\u003cli\u003eNdlovu N, Spillane C, McKeown PC, Cairns JE, Das B, Gowda M. Genome-wide association studies of grain yield and quality traits under optimum and low-nitrogen stress in tropical maize (Zea mays L.). Theor Appl Genet. 2022;135:4351\u0026ndash;70.\u003c/li\u003e\n\u003cli\u003eRaza A, Mubarik MS, Sharif R, Habib M, Jabeen W, Zhang C, et al. Developing drought-smart, ready-to-grow future crops. Plant Genome. 2023;16:e20279.\u003c/li\u003e\n\u003cli\u003eYuan Y, Cairns JE, Babu R, Gowda M, Makumbi D, Magorokosho C, et al. Genome-wide association mapping and genomic prediction analyses reveal the genetic architecture of grain yield and flowering time under drought and heat stress conditions in maize. Front Plant Sci. 2019;9 January:1\u0026ndash;15.\u003c/li\u003e\n\u003cli\u003eMasuka B, Magorokosho C, Olsen M, Atlin GN, B\u0026auml;nziger M, Pixley K V., et al. Gains in maize genetic improvement in eastern and southern Africa: II. CIMMYT open-pollinated variety breeding pipeline. Crop Sci. 2017;57:180\u0026ndash;91.\u003c/li\u003e\n\u003cli\u003eFisher M, Abate T, Lunduka RW, Asnake W, Alemayehu Y, Madulu RB. Drought tolerant maize for farmer adaptation to drought in sub-Saharan Africa: Determinants of adoption in eastern and southern Africa. Clim Change. 2015;133:283\u0026ndash;99.\u003c/li\u003e\n\u003cli\u003ePrasanna BM, Cairns JE, Zaidi PH, Beyene Y, Makumbi D, Gowda M, et al. Beat the stress: breeding for climate resilience in maize for the tropical rainfed environments. Theor Appl Genet. 2021;134:1729\u0026ndash;52.\u003c/li\u003e\n\u003cli\u003eBrouwer, C. and Heibloem M. Irrigation Water Management: Irrigation Water Needs. Training Manual. Train Man No 3. 1980.\u003c/li\u003e\n\u003cli\u003eHe Z, Zhang T, Liu X, Shang X. Water-yield relationship responses of maize to ridge-furrow planting systems coupled with multiple irrigation levels in China\u0026rsquo;s Horqin Sandy land. Agronomy. 2018;8.\u003c/li\u003e\n\u003cli\u003eBadu-Apraku B, Fakorede MAB. Breeding Maize for Drought Tolerance. Springer, Cham; 2017.\u003c/li\u003e\n\u003cli\u003eMenkir A, Crossa J, Meseka S, Bossey B, Muhyideen O, Riberio PF, et al. Stacking Tolerance to Drought and Resistance to a Parasitic Weed in Tropical Hybrid Maize for Enhancing Resilience to Stress Combinations. Front Plant Sci. 2020;11:166.\u003c/li\u003e\n\u003cli\u003eAo S, Russelle MP, Varga T, Feyereisen GW, Coulter JA. Drought tolerance in maize is influenced by timing of drought stress initiation. Crop Sci. 2020;60:1591\u0026ndash;606.\u003c/li\u003e\n\u003cli\u003eEdmeades, G. O., Trevisan, W., Prasanna, B. M., \u0026amp; Campos H. Tropical Maize (Zea mays L.).In:Genetic Improvement of Tropical Crops. Springer Cham; 2017.\u003c/li\u003e\n\u003cli\u003eMenkir A, Dieng I, Gedil M, Mengesha W, Oyekunle M, Riberio PF, et al. Approaches and progress in breeding drought-tolerant maize hybrids for tropical lowlands in west and central Africa. Plant Genome. 2024;17.\u003c/li\u003e\n\u003cli\u003eMcMillen MS, Mahama AA, Sibiya J, L\u0026uuml;bberstedt T, Suza WP. Improving drought tolerance in maize: Tools and techniques. Frontiers in Genetics. 2022;13 October:1\u0026ndash;13.\u003c/li\u003e\n\u003cli\u003eB\u0026auml;nziger M, Edmeades GO, Beck D, Bellon M. Breeding for Drought and Nitrogen Stress Tolerance in Maize: From Theory to Practice. Mex DF CIMMYT. 2000;:68.\u003c/li\u003e\n\u003cli\u003eCooper M, Gho C, Leafgren R, Tang T, Messina C. Breeding drought-tolerant maize hybrids for the US corn-belt: Discovery to product. Journal of Experimental Botany. 2014;65:6191\u0026ndash;4.\u003c/li\u003e\n\u003cli\u003eCooper M, Messina CD. Breeding crops for drought-Affected environments and improved climate resilience. Plant Cell. 2023;35:162\u0026ndash;86.\u003c/li\u003e\n\u003cli\u003eWang N, Liu B, Liang X, Zhou Y, Song J, Yang J, et al. Genome-wide association study and genomic prediction analyses of drought stress tolerance in China in a collection of off-PVP maize inbred lines. Mol Breed. 2019;39.\u003c/li\u003e\n\u003cli\u003eVarshney RK, Barmukh R, Bentley A, Nguyen HT. Exploring the genomics of abiotic stress tolerance and crop resilience to climate change. Plant Genome. 2024; February:1\u0026ndash;6.\u003c/li\u003e\n\u003cli\u003eLi C, Sun B, Li Y, Liu C, Wu X, Zhang D, et al. Numerous genetic loci identified for drought tolerance in the maize nested association mapping populations. BMC Genomics. 2016;17:1\u0026ndash;11.\u003c/li\u003e\n\u003cli\u003eLiu S, Qin F. Genetic dissection of maize drought tolerance for trait improvement. In: Molecular Breeding. 2021. p. 8.\u003c/li\u003e\n\u003cli\u003eShi W, Hao C, Zhang Y, Cheng J, Zhang Z, Liu J, et al. A combined association mapping and linkage analysis of kernel number per spike in common wheat (Triticum aestivum L.). Front Plant Sci. 2017;8 August:1\u0026ndash;13.\u003c/li\u003e\n\u003cli\u003eDavoud Torkamaneh FB. Genome-Wide Association Studies. 1st edition. New York, NY: Humana New York, NY; 2022.\u003c/li\u003e\n\u003cli\u003eAlseekh S, Kostova D, Bulut M, Fernie AR. Genome-wide association studies: assessing trait characteristics in model and crop plants. Cellular and Molecular Life Sciences. 2021;78:5743\u0026ndash;54.\u003c/li\u003e\n\u003cli\u003eXiao Y, Liu H, Wu L, Warburton M, Yan J. Genome-wide Association Studies in Maize: Praise and Stargaze. Mol Plant. 2017;10:359\u0026ndash;74.\u003c/li\u003e\n\u003cli\u003eTibbs Cortes L, Zhang Z, Yu J. Status and prospects of genome-wide association studies in plants. Plant Genome. 2021;14.\u003c/li\u003e\n\u003cli\u003eHao D, Xue L, Zhang Z, Cheng Y, Chen G, Zhou G, et al. Combined linkage and association mapping reveal candidate loci for kernel size and weight in maize. Breed Sci. 2019;69:420\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eZheng H, Chen J, Mu C, Makumbi D, Xu Y, Mahuku G. Combined linkage and association mapping reveal QTL for host plant resistance to common rust (Puccinia sorghi) in tropical maize. BMC Plant Biol. 2018;18:1\u0026ndash;14.\u003c/li\u003e\n\u003cli\u003eAn Y, Chen L, Li YX, Li C, Shi Y, Zhang D, et al. Genome-wide association studies and whole-genome prediction reveal the genetic architecture of KRN in maize. BMC Plant Biol. 2020;20:1\u0026ndash;11.\u003c/li\u003e\n\u003cli\u003ePrice AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, Reich D. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet. 2006;38:904\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eZhang Z, Ersoz E, Lai CQ, Todhunter RJ, Tiwari HK, Gore MA, et al. Mixed linear model approach adapted for genome-wide association studies. Nat Genet. 2010;42:355\u0026ndash;60.\u003c/li\u003e\n\u003cli\u003eLi M, Liu X, Bradbury P, Yu J, Zhang YM, Todhunter RJ, et al. Enrichment of statistical power for genome-wide association studies. BMC Biol. 2014;12.\u003c/li\u003e\n\u003cli\u003eKang HM, Sul JH, Service SK, Zaitlen NA, Kong SY, Freimer NB, et al. Variance component model to account for sample structure in genome-wide association studies. Nat Genet. 2010;42:348\u0026ndash;54.\u003c/li\u003e\n\u003cli\u003eZhou X, Stephens M. Genome-wide efficient mixed-model analysis for association studies. Nat Genet. 2012;44:821\u0026ndash;4.\u003c/li\u003e\n\u003cli\u003eYoosefzadeh-Najafabadi M, Eskandari M, Belzile F, Torkamaneh D. Genome-Wide Association Study Statistical Models: A Review. In: Methods in Molecular Biology. 2022. p. 43\u0026ndash;62.\u003c/li\u003e\n\u003cli\u003eZhang YW, Tamba CL, Wen YJ, Li P, Ren WL, Ni YL, et al. mrMLM v4.0.2: An R Platform for Multi-locus Genome-wide Association Studies. Genomics, Proteomics Bioinforma. 2020;18:481\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eZhang YM, Jia Z, Dunwell JM. Editorial: The applications of new multi-locus GWAS methodologies in the genetic dissection of complex traits, volume II. Front Plant Sci. 2023;14 December:1\u0026ndash;3.\u003c/li\u003e\n\u003cli\u003eLiu, X., Huang, M., Fan, B., Buckler, E. S., \u0026amp; Zhang Z. Iterative Usage of Fixed and Random Effect Models for Powerful and Efficient Genome- Wide Association Studies. PLOS Genet. 2016;12:1\u0026ndash;24.\u003c/li\u003e\n\u003cli\u003eHuang M, Liu X, Zhou Y, Summers RM, Zhang Z. BLINK: A package for the next level of genome-wide association studies with both individuals and markers in the millions. Gigascience. 2019;8:1\u0026ndash;12.\u003c/li\u003e\n\u003cli\u003eWang SB, Feng JY, Ren WL, Huang B, Zhou L, Wen YJ, et al. Improving power and accuracy of genome-wide association studies via a multi-locus mixed linear model methodology. Sci Rep. 2016;6.\u003c/li\u003e\n\u003cli\u003eTamba CL, Zhang Y-M. A fast mrMLM algorithm for multi-locus genome-wide association studies. bioRxiv. 2018.\u003c/li\u003e\n\u003cli\u003eWen YJ, Zhang H, Ni YL, Huang B, Zhang J, Feng JY, et al. Methodological implementation of mixed linear models in multi-locus genome-wide association studies. Brief Bioinform. 2018;19:700\u0026ndash;12.\u003c/li\u003e\n\u003cli\u003eZhang J, Feng JY, Ni YL, Wen YJ, Niu Y, Tamba CL, et al. PLARmEB: Integration of least angle regression with empirical Bayes for multilocus genome-wide association studies. Heredity (Edinb). 2017;118:517\u0026ndash;24.\u003c/li\u003e\n\u003cli\u003eRen WL, Wen YJ, Dunwell JM, Zhang YM. PKWmEB: Integration of Kruskal-Wallis test with empirical Bayes under polygenic background control for multi-locus genome-wide association study. Heredity (Edinb). 2018;120:208\u0026ndash;18.\u003c/li\u003e\n\u003cli\u003eTamba CL, Ni YL, Zhang YM. Iterative sure independence screening EM-Bayesian LASSO algorithm for multi-locus genome-wide association studies. PLoS Comput Biol. 2017;13.\u003c/li\u003e\n\u003cli\u003eKhan N, Shazadee H, Cloutier S, You FM. Genomics Assisted Breeding Strategy in Flax. In: You FM, Fofana B, editors. The Flax Genome. Compendium of Plant Genomes. 1st edition. Springer, Cham; 2023. p. 253\u0026ndash;72.\u003c/li\u003e\n\u003cli\u003eMeuwissen THE, Hayes BJ, Goddard ME. Prediction of total genetic value using genome-wide dense marker maps. Genetics. 2001;157:1819\u0026ndash;29.\u003c/li\u003e\n\u003cli\u003eCrossa J, P\u0026eacute;rez-Rodr\u0026iacute;guez P, Cuevas J, Montesinos-L\u0026oacute;pez O, Jarqu\u0026iacute;n D, de los Campos G, et al. Genomic Selection in Plant Breeding: Methods, Models, and Perspectives. Trends in Plant Science. 2017;22:961\u0026ndash;75.\u003c/li\u003e\n\u003cli\u003eBeyene Y, Semagn K, Mugo S, Tarekegne A, Babu R, Meisel B, et al. Genetic gains in grain yield through genomic selection in eight bi-parental maize populations under drought stress. Crop Sci. 2015;55:154\u0026ndash;63.\u003c/li\u003e\n\u003cli\u003eWallace JG, Zhang X, Beyene Y, Semagn K, Olsen M, Prasanna BM, et al. Genome-wide association for plant height and flowering time across 15 tropical maize populations under managed drought stress and well-watered conditions in Sub-Saharan Africa. Crop Sci. 2016;56:2365\u0026ndash;78.\u003c/li\u003e\n\u003cli\u003eZhang A, Wang H, Beyene Y, Semagn K, Liu Y, Cao S, et al. Effect of trait heritability, training population size and marker density on genomic prediction accuracy estimation in 22 bi-parental tropical maize populations. Front Plant Sci. 2017;8:1916.\u003c/li\u003e\n\u003cli\u003eZhang X, P\u0026eacute;rez-Rodr\u0026iacute;guez P, Semagn K, Beyene Y, Babu R, L\u0026oacute;pez-Cruz MA, et al. Genomic prediction in biparental tropical maize populations in water-stressed and well-watered environments using low-density and GBS SNPs. Heredity (Edinb). 2015;114:291\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eBeyene Y, Gowda M, P\u0026eacute;rez-Rodr\u0026iacute;guez P, Olsen M, Robbins KR, Burgue\u0026ntilde;o J, et al. Application of Genomic Selection at the Early Stage of Breeding Pipeline in Tropical Maize. Front Plant Sci. 2021;12:1265.\u003c/li\u003e\n\u003cli\u003eMasuka B, Araus JL, Das B, Sonder K, Cairns JE. Phenotyping for Abiotic Stress Tolerance in Maize. Journal of Integrative Plant Biology. 2012;54:238\u0026ndash;49.\u003c/li\u003e\n\u003cli\u003eZaidi PH. Management of drought stress in field phenotyping. Mexico City Mexico: CIMMYT; 2019.\u003c/li\u003e\n\u003cli\u003eAlvarado G, Rodr\u0026iacute;guez FM, Pacheco A, Burgue\u0026ntilde;o J, Crossa J, Vargas M, et al. META-R: A software to analyze data from multi-environment plant breeding trials. Crop J. 2020;8:745\u0026ndash;56.\u003c/li\u003e\n\u003cli\u003eBates D, M\u0026auml;chler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67.\u003c/li\u003e\n\u003cli\u003eKresovich S. Quantitative genetics in maize breeding. F Crop Res. 1990;23:78\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eSemagn K. Leaf tissue sampling and DNA extraction protocols. Methods Mol Biol. 2014;1115:53\u0026ndash;67.\u003c/li\u003e\n\u003cli\u003eElshire RJ, Glaubitz JC, Sun Q, Poland JA, Kawamoto K, Buckler ES, et al. A robust, simple genotyping-by-sequencing (GBS) approach for high diversity species. PLoS One. 2011;6:19379.\u003c/li\u003e\n\u003cli\u003eGlaubitz JC, Casstevens TM, Lu F, Harriman J, Elshire RJ, Sun Q, et al. TASSEL-GBS: A high capacity genotyping by sequencing analysis pipeline. PLoS One. 2014;9.\u003c/li\u003e\n\u003cli\u003eBradbury PJ, Zhang Z, Kroon DE, Casstevens TM, Ramdoss Y, Buckler ES. TASSEL: Software for association mapping of complex traits in diverse samples. Bioinformatics. 2007;23:2633\u0026ndash;5.\u003c/li\u003e\n\u003cli\u003eHubisz MJ, Falush D, Stephens M, Pritchard JK. Inferring weak population structure with the assistance of sample group information. Mol Ecol Resour. 2009;9:1322\u0026ndash;32.\u003c/li\u003e\n\u003cli\u003ePritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000;155:945\u0026ndash;59.\u003c/li\u003e\n\u003cli\u003eEarl DA, vonHoldt BM. STRUCTURE HARVESTER: A website and program for visualizing STRUCTURE output and implementing the Evanno method. Conserv Genet Resour. 2012;4:359\u0026ndash;61.\u003c/li\u003e\n\u003cli\u003eEvanno G, Regnaut S, Goudet J. Detecting the number of clusters of individuals using the software STRUCTURE: A simulation study. Mol Ecol. 2005;14:2611\u0026ndash;20.\u003c/li\u003e\n\u003cli\u003eWang J, Zhang Z. GAPIT Version 3: Boosting Power and Accuracy for Genomic Association and Prediction. Genomics, Proteomics Bioinforma. 2021;19:629\u0026ndash;40.\u003c/li\u003e\n\u003cli\u003eLipka AE, Tian F, Wang Q, Peiffer J, Li M, Bradbury PJ, et al. GAPIT: Genome association and prediction integrated tool. Bioinformatics. 2012;28:2397\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eVanRaden PM. Efficient methods to compute genomic predictions. J Dairy Sci. 2008;91:4414\u0026ndash;23.\u003c/li\u003e\n\u003cli\u003eHill WG, Weir BS. Variances and covariances of squared linkage disequilibria in finite populations. Theor Popul Biol. 1988;33:54\u0026ndash;78.\u003c/li\u003e\n\u003cli\u003eRemington DL, Thornsberry JM, Matsuoka Y, Wilson LM, Whitt SR, Doebley J, et al. Structure of linkage disequilibrium and phenotypic associations in the maize genome. Proc Natl Acad Sci U S A. 2001;98:11479\u0026ndash;84.\u003c/li\u003e\n\u003cli\u003eLi MX, Yeung JMY, Cherny SS, Sham PC. Evaluating the effective numbers of independent tests and significant p-value thresholds in commercial genotyping arrays and public imputation reference datasets. Hum Genet. 2012;131:747\u0026ndash;56.\u003c/li\u003e\n\u003cli\u003eGao X, Starmer \u0026Atilde;J, Martin ER. A Multiple Testing Correction Method for Genetic Association Studies Using Correlated Single Nucleotide Polymorphisms. Genet Epedemiology. 2008;369:361\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eSlaten ML, Chan YO, Shrestha V, Lipka AE, Angelovici R. HAPPI GWAS: Holistic analysis with pre- And post-integration GWAS. Bioinformatics. 2020;36:4655\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eJiao Y, Peluso P, Shi J, Liang T, Stitzer MC, Wang B, et al. Improved maize reference genome with single-molecule technologies Complete and accurate reference genomes and annotations provide fundamental tools for characterization of genetic and functional variation. Nat Publ Gr. 2017.\u003c/li\u003e\n\u003cli\u003eSchnable PS, Ware D, Fulton RS, Stein JC, Wei F, Pasternak S, et al. The B73 maize genome: Complexity, diversity, and dynamics. Science (80- ). 2009;326:1112\u0026ndash;5.\u003c/li\u003e\n\u003cli\u003eDanecek P, Auton A, Abecasis G, Albers CA, Banks E, DePristo MA, et al. The variant call format and VCFtools. Bioinformatics. 2011;27:2156\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eBarrett JC, Fry B, Maller J, Daly MJ. Haploview: Analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21:263\u0026ndash;5.\u003c/li\u003e\n\u003cli\u003eZhang R, Jia G, Diao X. geneHapR: an R package for gene haplotypic statistics and visualization. BMC Bioinformatics. 2023;24.\u003c/li\u003e\n\u003cli\u003eGabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, et al. The structure of haplotype blocks in the human genome. Science (80- ). 2002;296:2225\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eEndelman JB. Ridge Regression and Other Kernels for Genomic Selection with R Package rrBLUP. Plant Genome. 2011;4:250\u0026ndash;5.\u003c/li\u003e\n\u003cli\u003eWickham H. ggplot2. Springer-V. New York, NY: Springer New York; 2009.\u003c/li\u003e\n\u003cli\u003eCairns JE, Prasanna BM. Developing and deploying climate-resilient maize varieties in the developing world. Curr Opin Plant Biol. 2018;45:226\u0026ndash;30.\u003c/li\u003e\n\u003cli\u003eWang Y, Guo H, Wu X, Wang J, Li H, Zhang R. Transcriptomic and physiological responses of contrasting maize genotypes to drought stress. Front Plant Sci. 2022;13.\u003c/li\u003e\n\u003cli\u003eBlancon J, Buet C, Dubreuil P, Tixier MH, Baret F, Praud S. Maize green leaf area index dynamics: genetic basis of a new secondary trait for grain yield in optimal and drought conditions. Theor Appl Genet. 2024;137:68.\u003c/li\u003e\n\u003cli\u003eCairns JE, Crossa J, Zaidi PH, Grudloyma P, Sanchez C, Luis Araus J, et al. Identification of drought, heat, and combined drought and heat tolerant donors in maize. Crop Sci. 2013;53:1335\u0026ndash;46.\u003c/li\u003e\n\u003cli\u003eTrachsel S, Leyva M, Lopez M, Suarez EA, Mendoza A, Montiel NG, et al. Identification of tropical maize germplasm with tolerance to drought, nitrogen deficiency, and combined heat and drought stresses. Crop Sci. 2016;56.\u003c/li\u003e\n\u003cli\u003eRezende WS, Beyene Y, Mugo S, Ndou E, Gowda M, Sserumaga JP, et al. Performance and yield stability of maize hybrids in stress-prone environments in eastern Africa. Crop J. 2020;8:107\u0026ndash;18.\u003c/li\u003e\n\u003cli\u003eBeyene Y, Mugo S, Semagn K, Asea G, Trevisan W, Tarekegne A, et al. Genetic distance among doubled haploid maize lines and their testcross performance under drought stress and non-stress conditions. Euphytica. 2013;192:379\u0026ndash;92.\u003c/li\u003e\n\u003cli\u003eChapman SC, Edmeades GO. Selection improves drought tolerance in tropical maize populations: II. Direct and correlated responses among secondary traits. Crop Sci. 1999;39:1315\u0026ndash;24.\u003c/li\u003e\n\u003cli\u003eNdlovu N, Gowda M, Beyene Y, Chaikam V, Nzuve FM, Makumbi D, et al. Genomic loci associated with grain yield under well-watered and water-stressed conditions in multiple bi-parental maize populations. Front Sustain Food Syst. 2024;8.\u003c/li\u003e\n\u003cli\u003eBola\u0026ntilde;os J, Edmeades GO. The importance of the anthesis-silking interval in breeding for drought tolerance in tropical maize. F Crop Res. 1996;48.\u003c/li\u003e\n\u003cli\u003eBadu-Apraku B, Akinwale RO, Franco J, Oyekunle M. Assessment of reliability of secondary traits in selecting for improved grain yield in drought and low-nitrogen environments. Crop Sci. 2012;52:2050\u0026ndash;62.\u003c/li\u003e\n\u003cli\u003eZhang X, Ren Z, Luo B, Zhong H, Ma P, Zhang H, et al. Genetic architecture of maize yield traits dissected by QTL mapping and GWAS in maize. Crop J. 2022;10:436\u0026ndash;46.\u003c/li\u003e\n\u003cli\u003eClark RM, Linton E, Messing J, Doebley JF. Pattern of diversity in the genomic region near the maize domestication gene tb1. Proc Natl Acad Sci U S A. 2004;101:700\u0026ndash;7.\u003c/li\u003e\n\u003cli\u003eRashid Z, Sofi M, Harlapur SI, Kachapur RM, Dar ZA, Singh PK, et al. Genome-wide association studies in tropical maize germplasm reveal novel and known genomic regions for resistance to Northern corn leaf blight. Sci Rep. 2020;10:1\u0026ndash;16.\u003c/li\u003e\n\u003cli\u003eLu Y, Shah T, Hao Z, Taba S, Zhang S, Gao S, et al. Comparative SNP and haplotype analysis reveals a higher genetic diversity and rapider LD decay in tropical than temperate germplasm in maize. PLoS One. 2011;6.\u003c/li\u003e\n\u003cli\u003eXiong X, Li J, Su P, Duan H, Sun L, Xu S, et al. Genetic dissection of maize (Zea mays L.) chlorophyll content using multi-locus genome-wide association studies. BMC Genomics. 2023;24.\u003c/li\u003e\n\u003cli\u003eZhou G, Zhu Q, Mao Y, Chen G, Xue L, Lu H, et al. Multi-Locus Genome-Wide Association Study and Genomic Selection of Kernel Moisture Content at the Harvest Stage in Maize. Front Plant Sci. 2021;12 July:1\u0026ndash;13.\u003c/li\u003e\n\u003cli\u003eWen YJ, Wu X, Wang S, Han L, Shen B, Wang Y, et al. Identification of QTN-by-environment interactions for yield related traits in maize under multiple abiotic stresses. Front Plant Sci. 2023;14 February.\u003c/li\u003e\n\u003cli\u003eGrotewold E, Gray J. Maize transcription factors. Handb Maize Genet Genomics. 2009;II:693\u0026ndash;713.\u003c/li\u003e\n\u003cli\u003eNingning Z, Binbin L, Fan Y, Jianzhong C, Yuqian Z, Yejian W, et al. Molecular mechanisms of drought resistance using genome-wide association mapping in maize (Zea mays L.). BMC Plant Biol. 2023;23:1\u0026ndash;17.\u003c/li\u003e\n\u003cli\u003eZhu Y, Liu Y, Zhou K, Tian C, Aslam M, Zhang B, et al. Overexpression of ZmEREBP60 enhances drought tolerance in maize. J Plant Physiol. 2022;275.\u003c/li\u003e\n\u003cli\u003eQi H, Liang K, Ke Y, Wang J, Yang P, Yu F, et al. Advances of Apetala2/Ethylene Response Factors in Regulating Development and Stress Response in Maize. International Journal of Molecular Sciences. 2023;24.\u003c/li\u003e\n\u003cli\u003eZhang J, Liao J, Ling Q, Xi Y, Qian Y. Genome-wide identification and expression profiling analysis of maize AP2/ERF superfamily genes reveal essential roles in abiotic stress tolerance. BMC Genomics. 2022;23.\u003c/li\u003e\n\u003cli\u003eFarmer LM, Book AJ, Lee KH, Lin YL, Fu H, Vierstraa RD. The RAD23 family provides an essential connection between the 26S proteasome and ubiquitylated proteins in Arabidopsis. Plant Cell. 2010;22:124\u0026ndash;42.\u003c/li\u003e\n\u003cli\u003eDoroodian P, Hua Z. The ubiquitin switch in plant stress response. Plants. 2021;10:1\u0026ndash;21.\u003c/li\u003e\n\u003cli\u003eWang X, An Y, Xu P, Xiao J. Functioning of PPR Proteins in Organelle RNA Metabolism and Chloroplast Biogenesis. Frontiers in Plant Science. 2021;12 February:1\u0026ndash;8.\u003c/li\u003e\n\u003cli\u003eXing H, Fu X, Yang C, Tang X, Guo L, Li C, et al. Genome-wide investigation of pentatricopeptide repeat gene family in poplar and their expression analysis in response to biotic and abiotic stresses. Sci Rep. 2018;8:1\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eZhu X, Ze M, Yin J, Chern M, Wang M, Zhang X, et al. A phosphofructokinase B-type carbohydrate kinase family protein, PFKB1, is essential for chloroplast development at early seedling stage in rice. Plant Sci. 2020;290.\u003c/li\u003e\n\u003cli\u003ePeng Li , Zhaoxia Li GX and JZ*. Trihelix Transcription Factor ZmThx20 Is Required for Kernel Development in Maize. Int J Mol Sci. 2021;22:12137.\u003c/li\u003e\n\u003cli\u003eZhao D, Gao F, Guan P, Gao J, Guo Z, Guo J, et al. Identification and analysis of differentially expressed trihelix genes in maize (Zea mays) under abiotic stresses. PeerJ. 2023;11.\u003c/li\u003e\n\u003cli\u003eRiggs JW, Rockwell NC, Cavales PC, Callis J. Identification of the plant ribokinase and discovery of a role for Arabidopsis Ribokinase in nucleoside metabolism. J Biol Chem. 2016;291:22572\u0026ndash;82.\u003c/li\u003e\n\u003cli\u003eYang J, Tang Z, Yang W, Huang Q, Wang Y, Huang M, et al. Genome-wide characterization and identification of Trihelix transcription factors and expression profiling in response to abiotic stresses in Chinese Willow (Salix matsudana Koidz). Front Plant Sci. 2023;14.\u003c/li\u003e\n\u003cli\u003eRan F, Wang Y, Jiang F, Yin X, Bi Y, Shaw RK, et al. Studies on Candidate Genes Related to Flowering Time in a Multiparent Population of Maize Derived from Tropical and Temperate Germplasm. Plants. 2024;13.\u003c/li\u003e\n\u003cli\u003eKhan SU, Zheng Y, Chachar Z, Zhang X, Zhou G, Zong N, et al. Dissection of Maize Drought Tolerance at the Flowering Stage Using Genome-Wide Association Studies. Genes (Basel). 2022;13.\u003c/li\u003e\n\u003cli\u003eLu Y, Zhang S, Shah T, Xie C, Hao Z, Li X, et al. Joint linkage-linkage disequilibrium mapping is a powerful approach to detecting quantitative trait loci underlying drought tolerance in maize. Proc Natl Acad Sci U S A. 2010;107:19585\u0026ndash;90.\u003c/li\u003e\n\u003cli\u003eRaunser S, Magnani R, Huang Z, Houtz RL, Trievel RC, Penczek PA, et al. Rubisco in complex with Rubisco large subunit methyltransferase. Proc Natl Acad Sci U S A. 2009;106:3160\u0026ndash;5.\u003c/li\u003e\n\u003cli\u003eMininno M, Brugi\u0026egrave;re S, Pautre V, Gilgen A, Ma S, Ferro M, et al. Characterization of chloroplastic fructose 1,6-bisphosphate aldolases as lysine-methylated proteins in plants. J Biol Chem. 2012;287:21034\u0026ndash;44.\u003c/li\u003e\n\u003cli\u003eLi J, Zhang Y, Gao Z, Xu X, Wang Y, Lin Y, et al. Plant U-box E3 ligases PUB25 and PUB26 control organ growth in Arabidopsis. New Phytol. 2021;229:403\u0026ndash;13.\u003c/li\u003e\n\u003cli\u003eTrujillo M. News from the PUB: Plant U-box type E3 ubiquitin ligases. Journal of Experimental Botany. 2018;69:371\u0026ndash;84.\u003c/li\u003e\n\u003cli\u003eLiu Y, Li C, Qin A, Deng W, Chen R, Yu H, et al. Genome-wide identification and transcriptome profiling expression analysis of the U-box E3 ubiquitin ligase gene family related to abiotic stress in maize (Zea mays L.). BMC Genomics. 2024;25.\u003c/li\u003e\n\u003cli\u003eJia T, Ge Q, Zhang S, Zhang Z, Liu A, Fan S, et al. UDP-Glucose Dehydrogenases: Identification, Expression, and Function Analyses in Upland Cotton (Gossypium hirsutum). Front Genet. 2021;11.\u003c/li\u003e\n\u003cli\u003eK\u0026auml;rk\u0026ouml;nen A, Murigneux A, Martinant JP, Pepey E, Tatout C, Dudley BJ, et al. UDP-glucose dehydrogenases of maize: A role in cell wall pentose biosynthesis. Biochem J. 2005;391:409\u0026ndash;15.\u003c/li\u003e\n\u003cli\u003eWu H, Bai B, Lu X, Li H. A gibberellin-deficient maize mutant exhibits altered plant height, stem strength and drought tolerance. Plant Cell Rep. 2023;42:1687\u0026ndash;99.\u003c/li\u003e\n\u003cli\u003eMart\u0026iacute;n-Trillo M, Cubas P. TCP genes: a family snapshot ten years later. Trends in Plant Science. 2010;15:31\u0026ndash;9.\u003c/li\u003e\n\u003cli\u003eLan T, Xiong W, Chen X, Mo B, Tang G. Plant cytoplasmic ribosomal proteins: an update on classification, nomenclature, evolution and resources. Plant J. 2022;110:292\u0026ndash;318.\u003c/li\u003e\n\u003cli\u003eLi Y, Liu F, Li P, Wang T, Zheng C, Hou B. An Arabidopsis Cytokinin-Modifying Glycosyltransferase UGT76C2 Improves Drought and Salt Tolerance in Rice. Front Plant Sci. 2020;11 November:1\u0026ndash;15.\u003c/li\u003e\n\u003cli\u003eJia H, Li M, Li W, Liu L, Jian Y, Yang Z, et al. A serine/threonine protein kinase encoding gene KERNEL NUMBER PER ROW6 regulates maize grain yield. Nat Commun. 2020;11.\u003c/li\u003e\n\u003cli\u003eShohat H, Eliaz NI, Weiss D. Gibberellin in tomato: metabolism, signaling and role in drought responses. Molecular Horticulture. 2021;1.\u003c/li\u003e\n\u003cli\u003eLi S. The Arabidopsis thaliana TCP transcription factors: A broadening horizon beyond development. Plant Signal Behav. 2015;10.\u003c/li\u003e\n\u003cli\u003eLiu Y, Chen Z, Zhang C, Guo J, Liu Q, Yin Y, et al. Gene editing of ZmGA20ox3 improves plant architecture and drought tolerance in maize. Plant Cell Rep. 2024;43.\u003c/li\u003e\n\u003cli\u003eJiao P, Liu T, Zhao C, Fei J, Guan S, Ma Y. ZmTCP14, a TCP transcription factor, modulates drought stress response in Zea mays L. Environ Exp Bot. 2023;208.\u003c/li\u003e\n\u003cli\u003eDing S, Cai Z, Du H, Wang H. Genome-wide analysis of TCP family genes in Zea mays l. Identified a role for ZmTCP42 in drought tolerance. Int J Mol Sci. 2019;20.\u003c/li\u003e\n\u003cli\u003eTakemiya A, Sugiyama N, Fujimoto H, Tsutsumi T, Yamauchi S, Hiyama A, et al. Phosphorylation of BLUS1 kinase by phototropins is a primary step in stomatal opening. Nat Commun. 2013;4.\u003c/li\u003e\n\u003cli\u003eZeiner A, Colina FJ, Citterico M, Wrzaczek M. CYSTEINE-RICH RECEPTOR-LIKE PROTEIN KINASES: their evolution, structure, and roles in stress response and development. Journal of Experimental Botany. 2023;74:4910\u0026ndash;27.\u003c/li\u003e\n\u003cli\u003eZhang Y, Tian H, Chen D, Zhang H, Sun M, Chen S, et al. Cysteine-rich receptor-like protein kinases: emerging regulators of plant stress responses. Trends in Plant Science. 2023;28:776\u0026ndash;94.\u003c/li\u003e\n\u003cli\u003eLi C, Chen G, Mishina K, Yamaji N, Ma JF, Yukuhiro F, et al. A GDSL-motif esterase/acyltransferase/lipase is responsible for leaf water retention in barley. Plant Direct. 2017;1.\u003c/li\u003e\n\u003cli\u003eShen G, Sun W, Chen Z, Shi L, Hong J, Shi J. Plant GDSL Esterases/Lipases: Evolutionary, Physiological and Molecular Functions in Plant Development. Plants. 2022;11.\u003c/li\u003e\n\u003cli\u003eSu HG, Zhang XH, Wang TT, Wei WL, Wang YX, Chen J, et al. Genome-Wide Identification, Evolution, and Expression of GDSL-Type Esterase/Lipase Gene Family in Soybean. Front Plant Sci. 2020;11.\u003c/li\u003e\n\u003cli\u003eLi Z, Ahammed GJ. Plant stress response and adaptation via anthocyanins: A review. Plant Stress. 2023;10.\u003c/li\u003e\n\u003cli\u003eLu N, Jun JH, Li Y, Dixon RA. An unconventional proanthocyanidin pathway in maize. Nat Commun. 2023;14.\u003c/li\u003e\n\u003cli\u003eZhang Q, Liu H, Wu X, Wang W. Identification of drought tolerant mechanisms in a drought-tolerant maize mutant based on physiological, biochemical and transcriptomic analyses. BMC Plant Biol. 2020;20:1\u0026ndash;14.\u003c/li\u003e\n\u003cli\u003eKushwaha HR, Joshi R, Pareek A, Singla-Pareek SL. MATH-domain family shows response toward abiotic stress in arabidopsis and rice. Front Plant Sci. 2016;7.\u003c/li\u003e\n\u003cli\u003eCai G, Zang Y, Wang Z, Liu S, Wang G. Arabidopsis BTB-A2s Play a Key Role in Drought Stress. Biology (Basel). 2024;13.\u003c/li\u003e\n\u003cli\u003eZeng T, Meng Z, Yue R, Lu S, Li W, Li W, et al. Genome wide association analysis for yield related traits in maize. BMC Plant Biol. 2022;22:1\u0026ndash;11.\u003c/li\u003e\n\u003cli\u003eGuo J, Sun B, He H, Zhang Y, Tian H, Wang B. Current understanding of bhlh transcription factors in plant abiotic stress tolerance. International Journal of Molecular Sciences. 2021;22.\u003c/li\u003e\n\u003cli\u003eZhang T, Lv W, Zhang H, Ma L, Li P, Ge L, et al. Genome-wide analysis of the basic Helix-Loop-Helix (bHLH) transcription factor family in maize. BMC Plant Biol. 2018;18.\u003c/li\u003e\n\u003cli\u003eNan GL, Teng C, Fernandes J, O\u0026rsquo;Connor L, Meyers BC, Walbot V. A cascade of bHLH-regulated pathways programs maize anther development. Plant Cell. 2022;34:1207\u0026ndash;25.\u003c/li\u003e\n\u003cli\u003eHe Z, Zhong J, Sun X, Wang B, Terzaghi W, Dai M. The maize ABA receptors ZmPYl8, 9, and 12 facilitate plant drought resistance. Front Plant Sci. 2018;9.\u003c/li\u003e\n\u003cli\u003eNiu L, Liu L, Wang W. Digging for Stress-Responsive Cell Wall Proteins for Developing Stress-Resistant Maize. Front Plant Sci. 2020;11.\u003c/li\u003e\n\u003cli\u003eSong N, Xu Z, Wang J, Qin Q, Jiang H, Si W, et al. Genome-wide analysis of maize CONSTANS-LIKE gene family and expression profiling under light/dark and abscisic acid treatment. Gene. 2018;673:1\u0026ndash;11.\u003c/li\u003e\n\u003cli\u003eWang F, Han T, Jeffrey Chen Z. Circadian and photoperiodic regulation of the vegetative to reproductive transition in plants. Commun Biol. 2024;7:1\u0026ndash;11.\u003c/li\u003e\n\u003cli\u003eBhat JA, Yu D, Bohra A, Ganie SA, Varshney RK. Features and applications of haplotypes in crop breeding. Commun Biol. 2021;4:1\u0026ndash;12.\u003c/li\u003e\n\u003cli\u003eBeyene Y, Gowda M, Olsen M, Robbins KR, P\u0026eacute;rez-Rodr\u0026iacute;guez P, Alvarado G, et al. Empirical Comparison of Tropical Maize Hybrids Selected Through Genomic and Phenotypic Selections. Front Plant Sci. 2019;10:1502.\u003c/li\u003e\n\u003cli\u003eErtiro BT, Labuschagne M, Olsen M, Das B, Prasanna BM, Gowda M. Genetic Dissection of Nitrogen Use Efficiency in Tropical Maize Through Genome-Wide Association and Genomic Prediction. Front Plant Sci. 2020;11 April:1\u0026ndash;16.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 2 are available in the Supplementary Files section\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-plant-biology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pbio","sideBox":"Learn more about [BMC Plant Biology](http://bmcplantbiol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pbio/default.aspx","title":"BMC Plant Biology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5289238/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5289238/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDrought is a major abiotic stress in sub-Saharan Africa, impacting maize growth and development leading to severe yield loss. Drought tolerance is a complex trait regulated by multiple genes, making direct grain yield selection ineffective. To dissect the genetic architecture of grain yield and flowering traits under drought stress, a genome-wide association study (GWAS) was conducted on a panel of 236 maize lines testcrossed and evaluated under managed drought and optimal growing conditions in multiple environments using seven multi-locus GWAS models (mrMLM, FASTmrMLM, FASTmrEMMA, pLARmEB, pKWmEB, ISIS EM-BLASSO, and FARMCPU) from \u003cem\u003emrMLM \u003c/em\u003eand \u003cem\u003eGAPIT R packages.\u003c/em\u003e Genomic prediction with RR-BLUP model was applied on BLUEs across locations under optimum and drought conditions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;A total of 172 stable and reliable quantitative trait nucleotides (QTNs) were identified, of which 77 are associated with GY, AD, SD, ASI, PH, EH, EPO and EPP under drought and 95 are linked to GY, AD, SD, ASI, PH, EH, EPO and EPP under optimal conditions, with 17 QTNs explaining over 10% of the phenotypic variation (\u003cem\u003eR\u003c/em\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e ≥ 10%). Furthermore, 43 candidate genes were discovered and annotated. Two major candidate genes, \u003cem\u003eZm00001eb041070\u003c/em\u003e closely associated with grain yield near peak QTN,\u003cem\u003e qGY_DS1.1\u003c/em\u003e (S1_216149215) and \u003cem\u003eZm00001eb364110\u003c/em\u003e closely related to anthesis-silking interval near peak QTN, \u003cem\u003eqASI_DS8.2\u003c/em\u003e (S8_167256316) were identified, encoding AP2-EREBP transcription factor 60 and TCP-transcription factor 20, respectively under drought stress. Haplo-pheno analysis identified superior haplotypes for \u003cem\u003eqGY_DS1.1\u003c/em\u003e (S1_216149215) associated with the higher grain yield under drought stress. Genomic prediction revealed moderate to high prediction accuracies under optimum and drought conditions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe lines carrying superior haplotypes can be used as potential donors in improving grain yield under drought stress. Integration of genomic selection with GWAS results leads not only to an increase in the prediction accuracy but also to validate the function of the identified candidate genes as well increase in the accumulation of favorable alleles with minor and major effects in elite breeding lines. This study provides valuable insight into the genetic architecture of grain yield and secondary traits under drought stress.\u003c/p\u003e","manuscriptTitle":"Genome-Wide Association Mapping and Genomic Prediction Analyses Reveal the Genetic Architecture of Grain Yield and Agronomic Traits under Drought and Optimum Conditions in Maize","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-25 08:19:17","doi":"10.21203/rs.3.rs-5289238/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-10-25T07:47:57+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-10-24T14:10:59+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-10-23T11:49:17+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Plant Biology","date":"2024-10-18T11:46:57+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-plant-biology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pbio","sideBox":"Learn more about [BMC Plant Biology](http://bmcplantbiol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pbio/default.aspx","title":"BMC Plant Biology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2ebf148c-f11b-4a9e-b803-02d52c3270d9","owner":[],"postedDate":"October 25th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-02-03T16:08:26+00:00","versionOfRecord":{"articleIdentity":"rs-5289238","link":"https://doi.org/10.1186/s12870-025-06135-3","journal":{"identity":"bmc-plant-biology","isVorOnly":false,"title":"BMC Plant Biology"},"publishedOn":"2025-02-01 15:57:16","publishedOnDateReadable":"February 1st, 2025"},"versionCreatedAt":"2024-10-25 08:19:17","video":"","vorDoi":"10.1186/s12870-025-06135-3","vorDoiUrl":"https://doi.org/10.1186/s12870-025-06135-3","workflowStages":[]},"version":"v1","identity":"rs-5289238","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5289238","identity":"rs-5289238","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}

Text is read by the "Ask this paper" AI Q&A widget below. Extraction quality varies by source — PMC NXML preserves structure cleanly, OA-HTML may include some navigation residue, and OA-PDF can have broken hyphenation. The publisher copy (via DOI) is the canonical version.

My notes (saved in your browser only)

⚙ Ask this paper AI returns verbatim quotes from the full text · source: preprint-html ⓘ

Answers must be backed by verbatim quotes from this paper's full text. Hallucinated quotes are dropped automatically; if no verbatim passage answers the question, we say so. How this works

Citation neighborhood (no data yet)

We don't have any in-corpus citations linked to this paper yet. This is a recent paper (2024) — citers typically take a year or two to land, and the OpenAlex reference graph may still be filling in.

Source provenance

europepmc: last seen: 2026-05-19T01:45:01.086888+00:00