Search

The efficiency of genomic selection strongly depends on the prediction accuracy of the genetic merit of candidates. Numerous papers have shown that the composition of the calibration set is a key contributor to prediction accuracy. A poorly defined calibration set can result in low accuracies, whereas an optimized one can considerably increase accuracy compared to random sampling, for a same size. Alternatively, optimizing the calibration set can be a way of decreasing the costs of phenotyping by enabling similar levels of accuracy compared to random sampling but with fewer phenotypic units. We present here the different factors that have to be considered when designing a calibration set, and review the different criteria proposed in the literature. We classified these criteria into two groups: model-free criteria based on relatedness, and criteria derived from the linear mixed model. We introduce criteria targeting specific prediction objectives including the prediction of highly diverse panels, biparental families, or hybrids. We also review different ways of updating the calibration set, and different procedures for optimizing phenotyping experimental designs.

Characterizing the genomic diversity of species is critical to understand the molecular origin of phenotypic variations. Whole genome sequence assemblies at the chromosome scale with low amount of missing data are critical resources for answering such questions. Our aim is to explore combination of technologies to answer those questions, and to find the best one. Using a specific Maize line, we have tested different sequencing applications to build a "high quality genome" and identify a large collection of genome variants. Here we demonstrate that the use of HIFI reads combined with Hi-C and Linked-reads generate a chromosome-scale genome assembly with a better contiguity than the reference genome B73. Moreover we show, that PacBio CLR reads allow a wide detection of structural variants of the genomes.

Understanding the mechanisms of adaptation to the environment in cultivated plants is a promising way to meet the challenge of maintaining food security in the context of global warming. In the case of maize, high-throughput sequencing has revealed that structural variations represent a large part of the genome and could have huge phenotypic effects. Among these, Presence Absence Variants (PAVs, which include insertion/deletion of large DNA sequences) may be involved in adaptation of maize to its environment, but their contribution to the genetic determinism of traits and genotype by environment interactions remains largely unknown. To address this issue, we performed a genome-wide association study between two types of polymorphisms, SNPs and InDels, and molecular traits obtained from proteomics and metabolomics analyses to detect quantitative trait loci (QTLs). The genetic panel used for this study was composed of 254 dent inbred lines genotyped with 978,134 SNPs and 72,041 InDels. The latter encompassed from 37 to 129,700 pb, including thousands of PAVs that are not present in the B73 reference genome. Proteins and metabolites were quantified by mass spectrometry in leaf samples from F1 hybrids obtained by crossing the inbred lines with one flint tester line. Hybrid plants were grown under two watering conditions (well-watered and water deficit) in greenhouse. In total, we detected 61,225 QTLs associated with proteome or metabolome variations. Among these, 4,766 QTLs were exclusively detected by InDels. To take into account the difference of marker density between InDels and SNPs, we used a re-sampling approach which showed that there was no difference between InDels and SNPs regarding the number and effect of the QTLs detected . Additionally, the QTLs detected by the two types of polymorphism were equally distributed in the two watering conditions. These preliminary results show that InDels are worse considering to detect new genetic regions of interest. They also suggest that InDels and SNPs equally contribute to molecular trait variation and response to drought stress.

International audience; Aquaporins are channel proteins facilitating the movement of water and other small solutes across biological membranes. They are involved in multiple cellular and physiological processes and play indisputable roles in the adaptation to the environment. While the dynamics of aquaporin genes expression profiles over daytime, organs, development stages and in response to environmental cues, has been extensively reported, descriptions of some precise molecular components of this regulation are scarce.Measurements of gene expression of five plasma membrane aquaporins (PIPgenes) in maize leaves (elongation and mature zones) across a 252-hybrid panel (DROPS panel) by RT-qPCR revealed the existing diversity and pointed at different profiles of variability thatimply different effects of selection over time on different aquaporin isoforms. GWAS and covariate GWAS allowed mapping numerous eQTLs, both local and distant. The implementation of a double fluorescent promoter activity reporter assay, transformed by biolistic into entire maize plants, allowed the evaluation of bothcisandtrans-acting regulatory candidates. Among them, a MITE-containing indel within aPIPgene promoter region, and a distant NAC transcription factor, were successfully identified as actors of thePIPgene regulation. Further steps should be directed to reveal the physiological relevance of such regulations.

International audience; Understanding the mechanisms of adaptation to the environment in cultivated plants is a promising way to meet the challenge of maintaining food security in the context of global warming. In the case of maize, high-throughput sequencing has revealed that structural variations represent a large part of the genome and could have huge phenotypic effects. Among these, Presence Absence Variants (PAVs) may be involved in adaptation of maize to its environment, but their contribution to the genetic determinism of traits and genotype by environment interactions remains largely unknown. To address this issue, we performed a genome-wide association study between two types of polymorphisms, SNPs and Insetions/Deletions (InDels), and molecular traits obtained from proteomics and metabolomics analyses. The genetic panel used for this study was composed of 254 dent inbred lines genotyped with 978,134 SNPs and 72,041 InDels. The latter encompassed from 37 to 129,700 pb, including thousands of PAVs that are not present in the B73 reference genome. Proteins and metabolites were quantified by mass spectrometry in leaf samples from F1 hybrids obtained by crossing the inbred lines with one flint tester line. Hybrid plants were grown under two watering conditions (well-watered and water deficit) in greenhouse. In total, we detected 61,225 QTLs associated with proteome or metabolome variations. Among these, 4,766 QTLs were exclusively detected by InDels. To take into account the difference of marker density between InDels and SNPs, we used a re-sampling approach which showed that there is no difference for effect size distribution of QTLs between InDels and SNPs and for the number of QTLs detected by InDels or SNPs. Additionally, the QTLs detected by the two types of polymorphism were equally distributed in the two watering conditions. Our results suggest that InDels and SNPs equally contributed to molecular trait variation and response to drought stress.

From the 17th century until the arrival of hybrids in 1960s, maize landraces were cultivated in the South-West of France, a traditional region for maize cultivation. A set of landraces were collected in this region between the 1950s and 1980s and were then conservedex situin a germplam collection. Previous studies using molecular markers on approx. twenty landraces fo this region showed that they belonged to a Pyrenees-Galicia Flint genetic group and originated from hybridization between Caribbean and Northern Flint germplasms introduced in Europe. In this study, we assessed the structure and genetic diversity of 194 SWF maize landraces to elucidate their origin, using a 50K SNP array and a bulk DNA approach. We identified two weakly differentiated genetic groups, one in the Western part and the other in the Eastern part. We highlighted the existence of a longitudinal gradient along the SWF area that was probably maintained through the interplay between genetic drifts and restricted gene flows, rather than through differential climatic adaptation. The contact zone between the two groups observed near the Garonne valley may be the result of these evolutionnary forces. We found only few significant cases of hybridization between Caribbean and Northern Flint germplasms in the region. We also found gene flows from various maize genetic groups to SWF landraces. Thus, we assumed that SWF landraces had a multiple origin with a slightly higher influence of Tropical germplasm in the West and preponderance of Northern Flint germplasm in the East.

Background and objectives Characterizing the genomic diversity of maize is critical to understand the molecular origin of structural variation, and is a prerequisite to underpin the functional variation underlying phenotypic variation. Whole genome sequence assemblies at the chromosome scale with low amount of missing data are critical resources for answering such questions. Whole genome sequence assemblies are now available for several maize inbred lines, but this is still missing for a number of key founders used in European maize breeding programs. This concerns the dent genetic pool and the European specific flint genetic group. Therefore, we have established a list of 13 priority lines for de novo sequencing as part of the SeqOccIn project.

We revisited, in a genomic context, the theory of hybrid genetic evaluation models of hybrid crosses of pure lines, as the current practice is largely based on infinitesimal model assumptions. Expressions for covariances between hybrids due to additive substitution effects and dominance and epistatic deviations were analytically derived. Using dense markers in a GBLUP analysis, it is possible to split specific combining ability into dominance and across-groups epistatic deviations, and to split general combining ability (GCA) into within-line additive effects and within-line additive by additive (and higher order) epistatic deviations. We analyzed a publicly available maize data set of Dent × Flint hybrids using our new model (called GCA-model) up to additive by additive epistasis. To model higher order interactions within GCAs, we also fitted “residual genetic” line effects. Our new GCA-model was compared with another genomic model which assumes a uniquely defined effect of genes across origins. Most variation in hybrids is accounted by GCA. Variances due to dominance and epistasis have similar magnitudes. Models based on defining effects either differently or identically across heterotic groups resulted in similar predictive abilities for hybrids. The currently used model inflates the estimated additive genetic variance. This is not important for hybrid predictions but has consequences for the breeding scheme—e.g. overestimation of the genetic gain within heterotic group. Therefore, we recommend using GCA-model, which is appropriate for genomic prediction and variance component estimation in hybrid crops using genomic data, and whose results can be practically interpreted and used for breeding purposes.

International audience

When handling a structured population in association mapping, group-specific allele effects may be observed at quantitative trait loci (QTLs) for several reasons: (i) a different linkage disequilibrium (LD) between SNPs and QTLs across groups, (ii) group-specific genetic mutations in QTL regions, and/or (iii) epistatic interactions between QTLs and other loci that have differentiated allele frequencies between groups. We present here a new genome-wide association (GWAS) approach to identify QTLs exhibiting such group-specific allele effects. We developed genetic materials including admixed progeny from different genetic groups with known genome-wide ancestries (local admixture). A dedicated statistical methodology was developed to analyze pure and admixed individuals jointly, allowing one to disentangle the factors causing the heterogeneity of allele effects across groups. This approach was applied to maize by developing an inbred “Flint-Dent” panel including admixed individuals that was evaluated for flowering time. Several associations were detected revealing a wide range of configurations of allele effects, both at known flowering QTLs (Vgt1, Vgt2 and Vgt3) and new loci. We found several QTLs whose effect depended on the group ancestry of alleles while others interacted with the genetic background. Our GWAS approach provides useful information on the stability of QTL effects across genetic groups and can be applied to a wide range of species., Author summary Identification of genomic regions involved in genetic architecture of traits has become commonplace in quantitative genetics studies. Genetic structure is a common feature in human, animal and plant species and most current methods target genomic regions whose effects on traits are conserved between genetic groups. However, a heterogeneity of allele effects may be observed due to different factors: a group-specific correlation between the alleles of the tagged marker and those of the causal variant, a group-specific mutation at the causal variant or an epistatic interaction between the causal variant and the genetic background. We propose a new method adapted to structured populations including admixed individuals, which aims to identify these genomic regions and to unravel the previous factors. The method was applied to a maize inbred diversity panel including lines from the dent and the flint genetic groups, as well as admixed lines, evaluated for flowering time. Several genomic regions were detected with various configurations of allele effects, with evidence of epistatic interactions between some of the loci and the genetic background.

International audience

International audience

Additional file 1. contains additional information on the simulation of genotypes, the simulation of phenotypes and the genomewide prediction model considered.

Additional file 2. details the usefulness criterion parental contributions based optimal cross selection methodology.

Le présent rapport rassemble les revues de la littérature scientifique internationale réalisées par le collectif d’experts scientifiques constitué dans le cadre de l’étude « Place des agricultures européennes dans le monde à l’horizon 2050 : entre enjeux climatiques et défis de la sécurité alimentaire » (ci-après désignée par l’acronyme AE2050).Pour rappel, l’étude AE2050 consiste à examiner, au moyen d’un modèle de bilans (GlobAgri-AE2050 ), l’évolution des surfaces, des niveaux de productions, et des échanges mondiaux de produits agricoles à l’horizon 2050 dans chacune des régions du monde , en fonction (i) de celle de la demande en produits agricoles (notamment liée à la croissance démographique et aux évolutions des régimes alimentaires) et (ii) des caractéristiques clefs des systèmes agricoles que sont les rendements (végétaux et animaux), les intensités culturales et les disponibilités en terres. On cherche, dans cette étude, à intégrer autant que possible les effets du changement climatique dans la projection des variables clefs décrivant les systèmes agricoles.

Additional file 4:Supplementary Figures contain the supplemental Figure S1. (Evolution of the additive genetic variance intra- and inter-family components in the breeding population); Figure S2. (Summary statistics on the introduction crosses); Figure S3. (Effect of TS composition on intra-family prediction accuracies); Figure S4. (Evolution of the breeding population over generations for two different weightings α).

International audience

National audience

National audience

The evolution of maize yields under drought is of particular concern in the context of climate change and human population growth. To better understand the mechanisms associated with the genetic polymorphisms underlying the variations of traits related to drought tolerance, we used a systems genetics approach integrating high-throughput phenotypic, proteomics and genomics data acquired on 254 maize hybrids grown under two watering conditions. We show that water deficit, even mild, induced a strong proteome remodeling and a reprogramming of the genetic control of the abundance of many proteins. We identify close co-localizations between QTLs and pQTLs, thus highlighting environment-specific pleiotropic loci associated to the co-expression of drought-responsive proteins and to the variations of phenotypic traits. These findings bring several lines of evidence supporting candidate genes at many loci and provide novel insight into the molecular mechanisms of drought tolerance.

International audience; European flint maize has proven to be an outstanding partner of North American dent maize for developing hybrids in Northern Europe. Use of publicly available genetic material in genetic studies is nevertheless complicated by a number of agronomic defaults, especially root and stalk lodging, tillering and a number of disease susceptibilities. A first series of lines available at INRA (Camus-Kulandaivelu et al. 2006, Genetics), was increased within the CornFed project, which involved German, Spanish and French partners, leading to 300 diverse lines. These lead to the discovery of QTL for biomass production (Rincent et al., 2014, Theor Appl Genet) and cold tolerance (Revilla et al., 2016, BMC plant Biology). To further increase this panel, Gouesnard et al. (2017, Theor appl Genet) conducted a GBS based survey of all flint type materials present in the INRA collection. Original lines not introgressed by dent germplasm were selected to complete the collection. They were also intercrossed to produce new recombinant materials between different genetic origins. This lead to a total of 1200 preexisting and new lines that were evaluated per se in a windy environment in Brittany to favor lodging, which revealed a high genetic variability. Association genetics will be conducted in the coming months. Finally, out of these lines, 350 lines selected within a narrow phenology range and showing limited lodging susceptibility have been crossed to a dent tester to conduct agronomic trials for adaptive traits, to be started in 2019.

Notes S2. GWAS statistical models and effects of confounding factors on GWAS. (DOCX 14 kb)

Figure S10. Examples of QTL detection on Chromosome 3, 6 and 8 for the different traits. The top panel represents the distribution of the QTLs along the chromosome of interest, for the different technologies. The vertical red line in this panel localizes the SNP chosen as reference for the QTL (marker with the strongest association). The middle panel is a zoom in the vicinity of the reference SNP, showing the Local distribution of the -log10(p-value). The bottom panel is the same zoom as the middle panel and shows the local linkage disequilibrium corrected by the kinship (r2k) of all SNPs, within this region, within the reference SNP. Ticks on different x-axes show the marker density of the three technologies (red for the 50K, blue for the 600K and green for the GBS, black for all markers). (DOCX 1632 kb)

Figure S8. QTL limits obtained by the LD_win approach projected on heatmaps representing the level of LD between associated SNPs for each trait (DTA: male flowering time, plantHT: plant height and GY: grain yield) and each chromosome. Upper and lower triangles on the heatmaps represented the r2 and r2K values between associated SNPs, respectively. Linkage disequilibrium between loci was colored according to values from weak LD (yellow) to high LD (red). The significant markers were ordered according to their physical positions on the chromosome and were represented by ticks on the four sides of the heatmaps. Limits of QTLs were displayed by gray dotted lines. QTL numbers were indicated in gray on the top and the right of each heatmap. (PDF 7360 kb)

Figure S7. QTL limits obtained by the LD_adj approach projected on heatmaps representing the level of LD between associated SNPs for each trait (DTA: male flowering time, plantHT: plant height and GY: grain yield) and each chromosome. Upper and lower triangles on the heatmaps represent the r2 and r2K values between associated SNPs, respectively. Linkage disequilibrium between loci was colored according to values from weak LD (yellow) to high LD (red). The significant markers were ordered according to their physical positions on the chromosome and were represented by ticks on the four sides of the heatmaps. Limits of QTLs were displayed by gray dotted lines. QTL numbers were indicated in gray on the top and the right of each heatmap. (PDF 7373 kb)

The evolution of maize yields under drought is of particular concern in the context of climate change and human population growth. To better understand the mechanisms associated with the genetic polymorphisms underlying the variations of traits related to drought tolerance, we used a systems genetics approach integrating high-throughput phenotypic, proteomics and genomics data acquired on 254 maize hybrids grown under two watering conditions. We show that water deficit, even mild, induced a strong proteome remodeling and a reprogramming of the genetic control of the abundance of many proteins. We identify close colocalizations between QTLs and pQTLs, thus highlighting environment-specific pleiotropic loci associated to the co-expression of drought-responsive proteins and to the variations of phenotypic traits. These findings bring several lines of evidence supporting candidate genes at many loci and provide novel insight into the molecular mechanisms of drought tolerance.

Figure S13. Colocalization of QTLs between the traits. Number of QTLs specific and shared by the three traits across all environments. Note that several QTLs from one trait were sometimes included in a single QTL of another trait. (DOCX 48 kb)

Figure S4. Contribution of four ancestral populations to 247 inbred lines after ADMIXTURE analysis. Markers from the 50K (top), 600K (middle) and GBS (bottom) were used. One vertical bar corresponds to one individual. Lines were ordered according to contributions observed for the 50K. From left to right, we have Stiff Stalk lines type B73 and B14a (blue), Iodent lines type PH207 (red), Lancaster lines type Mo17 and Oh43 (turquoise), a group of lines assembling W117, F7057 type lines (green). (DOCX 239 kb)

Figure S2. Comparison of genotyping data between 50K and 600K arrays, and GBS. (a) Distribution of minor allele frequency per SNP before filtering (monomorphic SNPs removed). (b) Distribution of SNP missing data proportion for the 50K array, 600K array, GBS direct reads (GBS1) and GBS after imputation by Cornell Institute (GBS2, note that the scale of the x-axes is different). (c) Relatedness distribution (Identity-By-State, IBS) after QC filtering with MAF≥1% (IBS using GBS1 was not estimated because of the low calling rate). (DOCX 71 kb)

Table S5. Narrow sense heritability (h2) and variance components (Vg, genetic variance; Ve, residual variance). The heritability and variance components were estimated for all traits (grain yield, male flowering time and plant height) using the R package Heritability [1]. (DOCX 18 kb)

When handling a structured population in association mapping, group-specific allele effects may be observed at quantitative trait loci (QTLs) for several reasons: (i) a different linkage disequilibrium (LD) between SNPs and QTLs across groups, (ii) group-specific genetic mutations in QTL regions, and/or (iii) epistatic interactions between QTLs and other loci that have differentiated allele frequencies between groups. We present here a new genome-wide association (GWAS) approach to identify QTLs exhibiting such group-specific allele effects. We developed genetic materials including admixed progeny from different genetic groups with known genome-wide ancestries (local admixture). A dedicated statistical methodology was developed to analyze pure and admixed individuals jointly, allowing one to disentangle the factors causing the heterogeneity of allele effects across groups. This approach was applied to maize by developing an inbred ”Flint-Dent” panel including admixed individuals that was evaluated for flowering time. Several associations were detected revealing a wide range of configurations of allele effects, both at known flowering QTLs (Vgt1, Vgt2 and Vgt3 ) and new loci. We found several QTLs whose effect depended on the group ancestry of alleles while others interacted with the genetic background. The existence of directional epistasis was highlighted by comparing admixed with pure individuals and was consistent with epistatic interactions identified at the level of QTLs. Our GWAS approach provides useful information on the stability of QTL effects across genetic groups and can be applied to a wide range of species.

Figure S9. Number of significant SNPs (blue line) and QTLs (red line) identified as a function of SNP density (x-axis) for the male flowering time (DTA), plant height (PlantHT) and grain yield (GY). (DOCX 125 kb)

Table S2. Stability of QTLs across environments for the male flowering time (DTA), Plant Height (PlantHT), Grain Yield (GY) and all traits. “Env. Nb” indicates the number of environment in which a QTL was detected. Next four columns indicate the number of QTL corresponding to each category. (DOCX 14 kb)

Figure S9. Number of significant SNPs (blue line) and QTLs (red line) identified as a function of SNP density (x-axis) for the male flowering time (DTA), plant height (PlantHT) and grain yield (GY). (DOCX 125 kb)

Figure S6. Heatmaps of genome-wide linkage disequilibrium (LD) between all markers within and between chromosomes using PANZEA SNPs from the 50K. All SNPs were ordered according to their position on the genome. “Unpl” after chromosome 10 refers to unplaced SNPs, in an arbitrary order. Dots represented LD between two loci and were colored according to their strength. Classical LD measurement r2 between loci were represented within triangle below the diagonal. Linkage disequilibrium corrected for structure (r2S, A), relatedness (r2K, B) or both (r2KS, C) were represented within triangle above the diagonal. (DOCX 442 kb)

Figure S3. Variation of the markers density, the recombination rate and the genome coverage in non-overlapping 2 Mbp windows along each chromosome except chromosome 3 (presented in Fig. 1). Markers have MAF above 5%. Top panel shows the variation of SNP number. In the bottom panel, dotted line represents the variation of recombination rate (cM / Mbp) and solid lines the proportion of genome covered by the SNPs using the cumulated length of physical LD windows around each SNP in each 2Mbp-windows. In these two panel, green, blue, red and black lines represent variation for GBS, 600K, 50K and combined technologies, respectively. Vertical dotted gray lines indicate limits of centromeric regions. Vertical lines between the two panels indicate the position of QTLs for flowering time (DTA), grain yield (GY) and Plant Height (PHT). Green, blue, red vertical lines indicate QTLs detected only by GBS, 600K and 50K technologies, respectively. Grey vertical lines indicate QTL detected by at least two technologies. Only QTL including a marker associated with -log10(pval) above 6 were shown. (PDF 152 kb)

The evolution of maize yields under drought is of particular concern in the context of climate change and human population growth. To better understand the mechanisms associated with the genetic polymorphisms underlying the variations of traits related to drought tolerance, we used a systems genetics approach integrating high-throughput phenotypic, proteomics and genomics data acquired on 254 half-sibs maize hybrids grown under well watered (WW) and mild water deficit (WD) conditions (Fig. 1).

Figure S5. Correlation between kinship matrix estimated by different technologies. Correlation (r) between the IBS and IBD (K_Freq) for each technology (A). Correlation of IBD (B) and IBS (D) between the three technologies (after imputation). (C) Correlation of IBD between the three technologies after removing the excess of rare alleles in the GBS to have the same distribution of MAF as in the 50K and the 600K. The red line is the bisector. (DOCX 255 kb)

The evolution of maize yields under drought is of particular concern in the context of climate change and human population growth. To better understand the mechanisms associated with the genetic polymorphisms underlying the variations of traits related to drought tolerance, we used a systems genetics approach integrating high-throughput phenotypic, proteomics and genomics data acquired on 254 maize hybrids grown under two watering conditions. We show that water deficit, even mild, induced a strong proteome remodeling and a reprogramming of the genetic control of the abundance of many proteins. We identify close colocalizations between QTLs and pQTLs, thus highlighting environment-specific pleiotropic loci associated to the co-expression of drought-responsive proteins and to the variations of phenotypic traits. These findings bring several lines of evidence supporting candidate genes at many loci and provide novel insight into the molecular mechanisms of drought tolerance.

Figure S11. Effect of minor allelic frequency distribution, SNP distributions along the genome and SNP densities on the number of associated SNP and QTL detected. Boxplot were drawn on 100 sets of 50 000 to 250 000 markers sampled according to different MAF distributions (A, B) and different SNP distributions along the genome (C, D). A, C: number of SNP associated; B, D: Number of QTL detected. In A and B, 600K_MAF (yellow), GBS_MAF (green), Low_MAF (cyan), Flat_MAF (blue), High_MAF (pink) on x axis indicate boxplots corresponding to MAF distribution similar to 600K, similar to GBS, skewed towards low MAF, flat MAF and skewed toward high MAF, respectively. In C and D, Dens_50K (red), Dens_600K (yellow), Dens_GBS (cyan), Dens_Gen (blue), Dens_Phys (pink) on x axis indicate distribution of SNPs along the genome corresponding to 50K, GBS, 600K, even genetic and physical distances, respectively. For A, B, C and D, modalities indicated as “Random” in x axis correspond to random sample of SNP. Number of markers for each boxplot are indicated after the point. (PDF 281 kb)

Figure S1. Different approaches used to impute missing data of the GBS. We considered the direct reads from GBS (GBS1) and four approaches for imputation (GBS2 to GBS5). GBS2 approach consisted in one imputation step from the direct read by Cornell University, using TASSEL software, but missing data was still present. GBS3 approach consisted in a genotype imputation of the whole missing data of the direct read by Beagle v3. In GBS4, genotype imputation by Beagle was performed on Cornell imputed data after replacing the heterozygous genotypes into missing data. GBS5, consisted in homozygous genotypes of GBS2 completed by values imputed in GBS3. (DOCX 14 kb)

Figure S8. QTL limits obtained by the LD_win approach projected on heatmaps representing the level of LD between associated SNPs for each trait (DTA: male flowering time, plantHT: plant height and GY: grain yield) and each chromosome. Upper and lower triangles on the heatmaps represented the r2 and r2K values between associated SNPs, respectively. Linkage disequilibrium between loci was colored according to values from weak LD (yellow) to high LD (red). The significant markers were ordered according to their physical positions on the chromosome and were represented by ticks on the four sides of the heatmaps. Limits of QTLs were displayed by gray dotted lines. QTL numbers were indicated in gray on the top and the right of each heatmap. (PDF 7360 kb)

Table S3. Proportion of low and high recombination regions, recombination rate and percentage of QTLs located in these regions for the three traits. “Chr” indicates the chromosome. Physical and genetic size columns indicated the size of each chromosome in bp and cM, respectively. Average recombination rate (“RecRate”) and proportion of the physical (“Phys”) and genetic (“Genetic) map in high recombination regions (“HighRec”, >0.5 cM / Mbp) for each chromosome are shown. Percentage of QTL in high recombination regions were displayed for three traits (DTA: male flowering, PlantHT: Plant Height, GY: Grain Yield). (DOCX 16 kb)

International audience

International audience

National audience; Maize landraces germplasm have a very large genetic diversity that is still poorly characterized and exploited in plant breeding programs. We studied the effect of both human selection and environmental adaptation on genome-wide diversity of landraces with a focus on landraces-hybrid transition in order to identify interesting source of genetic diversity to enlarge modern breeding pools. We developed a high-throughput, cheap and labor saving DNA pooling approach based on 50K SNP maize Illumina array and estimated thereby allelic frequencies of 23412 SNP in 156 landraces representing worldwide maize diversity. We compared the diversity of this collection at genome-wide scale level with that of a panel of 336 inbred lines. Our new approach: (i) gives accurate allelic frequencies estimation that are reproducible across laboratories, (ii) protects both against false detection of allele presence within landraces and against ascertainment bias. Modified Roger's genetic Distance estimated from 23412 SNP and 17 SSR on same DNA pool are highly correlated, which validates our approach. Accordingly, structuration analysis based on SNP gives consistent results with SSR for higher levels of structuration but gives a slightly different pictures for more advanced structuration levels, suggesting that SNP and SSR could capture differently recent evolution. Gene diversity of landraces varies strongly along the genome and according to geographic origins. We identified 376 SNP under diversifying selection unraveling a selective footprints in Tga1/Su1 regions. While some maize landraces were closely related to several inbred lines and strongly contributed to modern breeding pools as Reid Yellow Dent or Lancaster Surecrop, some other have no related inbred lines and seem to have poorly contributed. We identified limited diversity loss or selective sweep between landraces and inbred lines, excepted in centromeric regions. For these regions, original landraces could be interesting to enlarge genetic diversity of modern breeding pools.

International audience; In south-west of France, maize landraces had evolved under environmental condition and human management since 17th century and until the arrival of the hybrids in1960s. In the sixties, these landraces have been conserved ex situ at the maize biological resource center (INRA, Mauguio, France). Previous genetic studies of a sample of this collection allowed identifying a distinct genetic group named Pyrenees-Galicia. This group has been hypothesized to come from an hybridization between the Northern Flint group and the Caribbean group In this study, we analysed a broader sample of the INRA collection in order to determine their genetic structure and to describe their morphological diversity. Firstly, we analysed genetic diversity of 194 maize landraces from south-west of France with the 50K SNPs array and using a bulk DNA sample of 15 plants. A non-supervised admixture analysis was performed by adding 148 American and European landraces. This analysis shown that there were 8 genetic groups. In fact, two separate genetic groups were identified in South-west of France, one in the Western part and one in the Eastern part. Secondly, we assessed morphological differentiation between the two genetic groups found in south-west of France, using a principal component analysis and an analysis of variance on 15 traits. Landraces located in West part of south-western France are earlier with bigger kernels and ears with a lower number of rows than landraces located in East part. Finally, we performed a principal coordinate analysis (PCoA) on 194 maize landraces from south-west of France. We showed that landraces were distributed continuously along the first component of PCoA analysis. This component was correlated with geographical coordinates of the landrace collection sites, highlighting a longitudinal gradient and a latitudinal gradient.

Resumen del póster presentado en el Congreso celebrado en Saint-Malo, Francia, entre el 22 y el 25 de marzo de 2018., Genotyping-by-sequencing (GBS) is a highly cost-effective procedure that permits the analysis of large collections of inbred lines. We used it to characterize diversity in 1191 maize flint inbred lines from the INRA collection, the European Cornfed-Flint association panel, and flint lines recently derived from landraces. We analyzed the properties of GBS data obtained with different imputation methods, by comparison with a 50K SNP array. We identified 7 ancestral groups within the Flint collection (five typically flint: Northern Flint, Italy, Pyrenees-Galicia, Argentina, Lacaune, and also Dent and Pop corn) that are in agreement with breeding knowledge. This analysis highlighted that many lines are issued from crosses between different flint ancestral groups (admixture). Approximately 200 lines also appear to be issued from crosses with dent germplasm aiming at the improvement of flint germplasm. We performed association studies on different agronomic traits, revealing SNPs associated with cob color, kernel color, and male flowering time variation. We analyzed the relationship between the haplotype diversity and the trait variation at some strong association peaks., French National Research Agency (Amaizing, ANR-10-BTBR-03)

Size distribution of B73 and F2-specific sequences. All PAVs exhibit similar size distribution except F2 unanchored PAVs, which are more numerous and include larger sequences. (PDF 28 kb)