Genomic selection software

Mixed models have become a key tool for fitting genomic selection models, but most current genomic selection software can only include a single variance component other than the error, making hybrid prediction using additive, dominance and epistatic effects unfeasible for species displaying heterotic effects. Moreover, Likelihood-based software for fitting mixed models with multiple random effects that allows the user to specify the variance-covariance structure of random effects has not been fully exploited.

A new open-source R package called sommer is presented to facilitate the use of mixed models for genomic selection and hybrid prediction purposes using more than one variance component and allowing specification of covariance structures.

Lorenz et al. Yi and Jannink [ 22 ] suggest a multivariate approach for genomic selection of multiple traits to improve prediction accuracy on low heritability traits genetically correlated to high heritability traits.

There are some public efforts to build bioinformatics infrastructure for GS. This limitation creates challenges in a long-term storage, community access, analysis and data sharing. Also often, project-centric web-portals that use custom-designed database schemas are difficult to adapt to new projects.

Therefore, the solGS web application can be integrated easily into websites that use the ND database schema as backend for their data storage. The application can serve as a medium for community data and knowledge exchange, similar to the functioning of the SGN community annotation [ 29 ] and QTL analysis and linking to genomes tools [ 30 ]. Depending on data access privileges, solGS can facilitate web access and exchange of data on breeding material among a community of researchers.

Sharing GS output can be done conveniently through exchanging model output page links or data downloads. In the near future, we plan to integrate more features into the application to enhance the decision-making efficiency and capability of GS breeders. We will calculate superior progeny values of individuals based on expected mean values of progenies, expand the univariate RR-BLUP modeling into multivariate analysis, and run genetic correlation analysis and principal component analysis of individuals based on their genotypes.

Depending on the availability of R packages, we will add more modeling options such as the Bayesian methods and supervised classification algorithms.

We will add algorithms to preprocess phenotype data from experimental designs newly added to the database. We will write a comprehensive user manual and tutorials. To speed up the prediction process, we will parallelize analyses. It has an intuitive workflow for choosing a training population on which to fit a prediction model and estimating GEBVs of selection candidates.

Model input and output is visualized graphically and can be interactively explored or downloaded in text format. Its dependence on the generic, flexible, Chado ND database schema, for its data storage system, makes the tool adaptable to wide range of GS breeding programs. J Anim Breed Genet. Ann Bot. Brief Funct Genomics. PLoS One. Crop Sci. Heredity Edinb. Plant Genome. Article Google Scholar. G3 Bethesda. J Dairy Sci.

CAS Google Scholar. BMC Genet. Google Scholar. Database Oxford. Book Google Scholar. J Am Stat Assoc. Adv Agron. Jia Y, Jannink JL: Multiple-trait genomic selection methods increase genetic value prediction accuracy. ISMU 2. Nucleic Acids Res. Plant Physiol. BMC Bioinformatics. Download references. We thank to all members of the Cassava project who contributed in many ways.

We thank Suzy Strickler for proofreading the manuscript. We thank the anonymous reviewers for their constructive suggestions. You can also search for this author in PubMed Google Scholar. IYT conceptualized, designed and developed the algorithms, workflow and interface for the analysis and drafted the manuscript. JDE and NM wrote genotype and phenotype data loading scripts and loaded the data to the database.

JLJ provided scientific advice. LAM oversaw the development of the tool. All authors contributed in discussions and approved the final draft of the manuscript. The first method, shown in panel A, uses a trait name to search the database for individuals phenotyped for that trait and select the individuals from any number of trials Additional file 2. A second way panel B is to search for a training population or trials of interest and use the set of individuals evaluated in a trial or combination of trials.

PNG KB. Additional file 2: Example of a list of trials in all of which a given trait was phenotyped. When breeders search using a trait name for phenotyped individuals to create a training population and use in a prediction model, they get a list of relevant training populations or trials.

All Individuals from a trial or combination of trials can be used. PNG 59 KB. They can also study the phenotypic correlation among the traits Figure 3 A. PNG 77 KB. This article is published under license to BioMed Central Ltd.

Reprints and Permissions. Tecle, I. BMC Bioinformatics 15, Download citation. Received : 12 July Accepted : 26 November Published : 14 December Anyone you share the following link with will be able to read this content:. Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative. Skip to main content. Search all BMC articles Search. Download PDF. Abstract Background Genomic selection GS promises to improve accuracy in estimating breeding values and genetic gain for quantitative traits compared to traditional breeding methods.

Conclusions solGS enables breeders to store raw data and estimate GEBVs of individuals online, in an intuitive and interactive workflow. Background Genomic selection GS is a new breeding paradigm that promises higher accuracy in estimating breeding values and a higher rate of gain from selection per unit time for complex quantitative traits compared to phenotypic selection or traditional marker assisted selection MAS [ 1 ]-[ 3 ]. Implementation Software solGS is developed using open source software and runs on a Debian-based Linux server.

Figure 1. Maximize the efficiency of your breeding program by means of cutting edge data analysis technology wrapped in a higly intuitive and user-friendly web application. We assist breeding companies in all aspects of the breeding data life cycle, from the statistical design of experiments through data storage, analysis and decision support phases.

Get a hands-on experience using the software as a breeder of a virtual breeding company. Apply for a free trial login now. Progeno is a Ghent University spin-off company that aims to empower professional plant and animal breeders by giving them direct access to the state-of-the-art in breeding and selection methods by means of a user-friendly and industry-proven software system.

The Progeno software framework is centered around its highly innovative computing engine that allows to integrate all available breeding data, generally including many years of phenotypic trial observations and sometimes vast amounts of molecular marker information, into reliable genomic breeding values for all traits of interest. Making the best of your data, Progeno software warrants a more cost-effective breeding program as well as faster genetic progress which in turn are likely to secure a clear edge over your competition.

Please fill out the form below in case you would like to receive more information on the software or services that Progeno can provide. One of our team members will contact you as soon as possible! Skip to main content.