IMPORTANCE OF META-QTL ANALYSIS FOR CROP IMPROVEMENT

SHORT REVIEW PAPER

IMPORTANCE OF META-QTL ANALYSIS FOR CROP IMPROVEMENT
P. JANAKI RAMAYYA1 AND M. Y. DUDHE2
1

Senior Research Fellow & 2Scientist, Plant Breeding, Crop Improvement Section

Directorate of Oilseeds Research, Rajendranagar, Hyderabad 500030

ABSTRACT
In last few decades, large efforts have been made to produce crop genetic and genomic data at genome scale. A new challenge in breeding programs is to integrate information from genomics and that from QTL analysis, in order to identify sequences controlling the variation of important traits. Thus, discovering co-locations between candidate genes and QTLs is an essential step. Integrative QTL meta-analysis methods have been developed to take full advantage of existing results to more accurately predict the most probable location of QTL. Methods have been devised to build consensus map by integrating genetic map and QTL map positions and making it easier to search for co-localization between genes and QTL. Meta QTL is a suite of programs inbuilt in Biomercator software designed to carry out meta-analysis of QTL mapping experiments. The present study highlights the prerequisites, methodology for construction of consensus map, genetic maps for meta-QTL analysis and implications of this data in Marker Assisted Selection (MAS) for crop improvement.

INTRODUCTION TO QTL AND META-QTL ANALYSIS

There are number of reports on the mapping of quantitative trait loci (QTLs) in plants which has exponentially increased since the eighties, reaching a total of about 34,300 papers in 2011 (source: Google Scholar with key words QTL and plant). For species like maize and rice, this huge amount of QTL data has been recorded in databases (e.g. Gramene) that enable quick comparison of results obtained from the individual experiments. But for most species, QTL data accumulates in bibliography until the commencement of hot-spot genomic regions that become targets for introgression into breeding material or for cloning. To get a comprehensive understanding of the genetic control of a polygenic trait and to optimize its use in breeding, it is needed to get a complete view of the genetic architecture of the trait with the distribution of the involved loci along the genome. This synthesis can be greatly facilitated by achieving a QTL meta-analysis (Sarah Danan et al.2011). The general principle of a meta-analysis is to pool the results of several studies that address the same issue to improve the estimate of targeted parameters. Meta1

analysis was first used in social and medical sciences, like epidemiology. Recently, it was applied in plant genetics to combine on a single map, the genetic marker data and the QTL characteristics (location, confidence interval, effect and trait used for QTL detection) from independent QTL mapping experiments to estimate the optimal set of distinct consensus QTLs. The consistent QTL identified by meta-analysis for a set of QTLs at a confidence interval of 95% is called as meta-QTL (MQTL). The positions of those meta-QTLs are estimated with a higher accuracy as compared to the individual QTLs in the original experiments (Goffinet and Gerber 2000). To date, QTL meta-analyses have been achieved for traits related to plant development and plant response to environment (nutrients, abiotic and biotic stresses) in maize, wheat, rice, rapeseed, cotton, soybean, cocoa and apricot.

Agrobios Res., Vol. 1, No. 2. pp 121-124 (April-June, 2012)

P. JANAKI RAMAYYA et al.

META-QTL ANALYSIS, IMPORTANCE AND APPLICATIONS IN CROP IMPROVEMENT

Integrating QTL results from independent experiments performed on related species helps to survey the genetic diversity of loci/alleles underlying complex traits, and to highlight potential targets for breeding or QTL cloning. Meta-analysis helped to refine the genomic regions of interest frequently described, and provided the closest flanking markers (Sarah Danan et al.2011). QTL meta-analysis is an approach to identify consensus QTL across studies, to validate QTL effects across environments/genetic backgrounds and also to refine the QTL positions on the consensus map. Meta- QTL analysis requires independent QTLs for the same trait obtained from different plant populations, different locations, or different environmental conditions (Goffinet and Gerber 2000). The meta-QTL with smallest confidence interval (CI) and having consistent and large effect on the trait is useful in marker assisted selection (MAS). QTL regions harbor many genes; among them few key genes may be more important in the regulation of a complex trait. Meta-QTL regions with refined positions are more accurate for short listing candidate genes. The common candidate genes shortlisted across the metaQTLs are more likely candidates regulating the yield. Superior alleles of such key genes can also be mined from different sources and incorporated in elite cultivars to develop new varieties. Meta-analyses are designed to integrate information from diverse studies and conditions to evaluate treatment effects on a set of independent variables. Meta-analyses can be particularly informative because they provide a quantitative estimate of effect sizes. The effect size is calculated as the response to a specific treatment (e.g. irrigation) relative to a control (e.g. no irrigation). Although this approach evaluates quantitative treatments as qualitative (i.e. control vs. stressed), appropriate data partitioning captures the quantitative nature of treatment effects and can account for differences in treatment timing and intensity. The effect sizes for the response to nitrogen fertilization, for example, can be partitioned into dosage ranges. In some cases, however, it is not possible to parse treatments because their magnitude and intensity are not documented in comparable units. Nonetheless, judicious data prior to meta-analysis enables investigators to isolate potential experimental or biological influences on effect of size. Another application of meta-QTL analysis is for construction of consensus map and for genetic map, which is discussed in detail in the article.

PREREQUISITES AND SOFTWARE USED FOR META-QTL ANALYSIS

Statistical methods have been proposed for the metaanalysis of QTLs from several experiments. The metaanalysis was performed on the QTL clusters on each chromosome using the Biomercator 2.0 (Arcade et al.2004). The method proposed by Goffinet and Gerber (2000) was implemented in the Biomercator software. It compiles the QTLs that have been projected on an existing reference map and uses the transformed Akaike classification criterion to determine the best model between one QTL, two QTLs, three QTLs etc. until the maximum number of QTLs mapped in the same region. This method was first used by Chardon et al.(2004). Then Veyrieras et al.(2007) have extended the statistical method and implemented the new algorithms in the Meta QTL software. Meta QTL notably uses a weighted least squares strategy to build the consensus map from the maps of individual studies. It offers a new clustering approach based on a Gaussian mixture model to define the optimal number of QTL clusters or meta-QTLs on each chromosome that best explain the observed distribution of the individual projected QTLs. The Gaussian mixture model has shown to be flexible and robust to the non-independence of the experiments. Moreover, simulations demonstrated that the number of meta-QTLs selected by the Akaike criterion is lower than the expected number with random distributions of QTLs and that it has a very low probability to happen by chance. The Akaike Information Criterion (AIC) was used to select the QTL model on each chromosome (Akaike 1974). According to this criterion, the QTL model with least AIC value is considered the significant model indicating the number of meta-QTLs. Meta-QTL analysis requires a minimal set of descriptors characterizing each observed QTL: the QTL position, its confidence interval and/or its individual R value (at least one of them is mandatory), the trait related to the QTL and the size of the QTL mapping population used for the QTL detection. The statistical method implemented in the Biomercator software hypothesizes that the input mapping studies are independent from each other. QTL mapping studies, which were repeated in time and space, often detected redundant QTLs at the same position for the same trait.

METHODOLOGY FOR CONSTRUCTION OF CONSENSUS MAP AND GENETIC MAPS FOR META-ANALYSIS
The construction of the consensus map was performed chromosome by chromosome. To be able to

IMPORTANCE OF META-QTL ANALYSIS FOR CROP IMPROVEMENT

align the chromosome maps in the right orientation, a chromosome of a study should contain a minimum of two common markers with the corresponding chromosome of another study. QTL maps that did not share common markers were discarded from the construction of the consensus map. For several maps, few chromosomes were also missing, which lead to a variation of the number of input maps depending on the chromosome (Sarah Danan et al., 2011). Chromosomes connected with less than two common markers to the reference map were excluded before creation of the consensus map. In case of inversions of two markers between maps, only the marker that was present in the lowest number of maps was manually removed to ensure that the most frequent common markers would be systematically retained. The ConsMap command of the Meta QTL software version 1.0 was used to create the consensus marker map (Veyrieras et al., 2007). The implemented method is based on a Weighted Least Square (WLS) strategy, which made it possible to compile all the input maps into a consensus map in a single step. It takes into account the distances between adjacent markers from all individual maps rescaled in Haldane unit. The size and type of the mapping population are used to estimate the map accuracy and are integrated into the compilation. Marker names and positions were provided in the input map files, along with a file specifying the synonymous names of the same markers that were mapped in different maps. It is important to provide input genetic map information from different mapping experiments for the genetic map construction. For example, consider a set of n genetic mapping experiments concerning the same linkage group. These different experiments may involve different kinds of population pedigree. We consider that for each experiment i = 1,..., n only the estimated distances between ordered markers along the linkage group are available. We use ci, Ni, Mi to denote the populations cross design, the population size and the number or markers on the ith genetic map, respectively. Let's suppose that two markers aj and ak have been positioned on the ith map, ri,jk stands for the estimated recombination rate between markers a j and ak and di,jk = f [ri,jk] the corresponding estimated distance, where f is the mapping function which is assumed to be the same in the n mapping experiments (without loss of generality). Applying the classical asymptotic Gaussian distribution of the maximum-likelihood estimation of the parameter, we assume that the i,jk are normally distributed around the true recombination rate ri,jk between markers aj and ak with a variance var (ri,jk) = i2,jk. This variance i2,jk depends on the cross design ci, the value of ri,jk, the sample size Ni and the amount of information supplied by the marker pair aj and ak sampled population. Since
3

mapping functions are generally one to one functions, the functional invariance property of the maximumlikelihood estimate can be applied. Thus d i,jk is also normally distributed around the true distance denoted di,jk = f [ri,jk]. To obtain the variance of di,jk, denoted, r2i,jk.

FUTURE SCOPE AND CONCLUDING REMARK

The construction of consensus map requires the common set of markers among the linkage groups of the crop. In case of crop like sunflower ( Helianthus annuus L.) the maximum number of reports on QTL analysis (viz., yield, oil and 100-seed weight) are not available and whatever reports available, are based on the different molecular markers viz., RFLP, SSR and ISSR that means a common set of marker is lacking for the identified QTLs per study. Hence, it can be concluded that in crops like, sunflower, safflower, groundnut, castor, millets, sorghum, ckickpea, pigeonpea etc. where less QTLs were identified for the agronomically important traits, it will take time to go for the meta-QTL studies. However, in case of crops like rice and maize (where genome sequences is available) and maximum number of QTL reports are available for agronomically important traits with common markers across the linkage groups that can be utilized for the construction of a consensus map and for identification of meta-QTL regions. Hence, this technique can be utilized in crops where more number of studies for the QTL identification were carried out with common set of markers. In some of the studies, meta QTLs were used for deducing candidate genes. In case of wheat, meta-QTLs for Fusarium head blight resistance were recommended for marker-assisted selection after the completion of meta-analysis. It is revealed from the studies completed till now that the meta-analysis is useful in identifying consensus and precise QTLs for further development of crop improvement programmes by using MAS.

REFERENCES
Akaike H 1974. A new look at the statistical model identification. IEEE Trans Autom Control 19:716723. Arcade A, Labourdette A, Falque M, Mangin B, Chardon F, Charcosset A, Joets J 2004. Bio Mercator: integrating genetic maps and QTL towards discovery of candidate genes. Bioinformatics 20(14):2324-2326 Chardon F, Virlon B, Moreau L, Falque M, Joets J, Decousset L, Murigneux A, Charcosset A 2004. Genetic architecture of flowering time in maize as inferred from quantitative trait loci meta-analysis and synteny

Agrobios Res., Vol. 1, No. 2. pp 121-124 (April-June, 2012)

P. JANAKI RAMAYYA et al. conservation with rice genome. Genetics 168:2169 2185. Goffinet B, Gerber S 2000. Quantitative trait loci: a metaanalysis. Genetics 155:463473. Sarah Danan, Jean-Baptiste Veyrieras, Veronique Lefebvre 2011. Construction of a potato consensus map and QTL meta-analysis offer new insights into the genetic architecture of late blight resistance and plant maturity traits BMC Plant Biology 2011, 11:16. Veyrieras J B, Goffinet B, Charcosset A 2007. Meta QTL: a package of new computational methods for the metaanalysis of QTL Mapping experiments. BMC Bioinformatics, 8(1):49.