INTEGRATED VIEWING AND ANALYSIS OF PHENOTYPIC, GENOTYPIC, AND ENVIRONMENTAL DATA WITH "GENPHEN ARRAYS"

Authors: White, Jeffrey; HOOGENBOOM, GERRIT - UNIV OF GEORGIA

Submitted to: European Journal of Agronomy
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/15/2004
Publication Date: 9/7/2005
Citation: White, J.W., Hoogenboom, G. 2005. Integrated viewing and analysis of phenotypic, genotypic, and environmental data with "genphen arrays". European Journal of Agronomy. 23:170-182.

Interpretive Summary: A recurring challenge in agronomic research is how to interpret large data sets that combine information on the genetic makeup of the cultivars or breeder's lines being tested (the genotype), how traits are expressed in the field (the phenotype), and the locations or management conditions used to test the materials (the environment). High-resolution color graphics offer the possibility of presenting such data as pseudo-maps, where data are presented using dots or other symbols and an appropriate color scheme to indicate the range of values. This paper illustrates such 'GenPhEn arrays' (from genotype, phenotype and environment) using data from a multi-location wheat trial, which are used by researchers to identify high-yielding and stable lines that might be released as cultivars and from a study of common bean that uses a computer model to estimate how warmer temperatures might affect bean yields in Michigan. The arrays look somewhat like a colored map of point locations and allow researchers to view large amounts of data (i.e., over 5000 data points) in a compact and readily interpretable fashion and can be created with existing commercial software. Various modifications to the basic application are illustrated. It is expected that GenPhEn arrays can be used by plant breeders, agronomists, and others who are interested in rapidly examining large sets of data for patterns that merit further study using more quantitative approaches. This should increase the efficiency of data interpretation, ultimately leading to faster release of improved cultivars or of management recommendations. Simplified arrays might be useful for presenting research results to growers or other stakeholders.

Technical Abstract: A recurring challenge in agronomic research is how to interpret large data sets that combine information on genotypes, phenotypes, and environments. High resolution color graphics offer the possibility of presenting such data as pseudo-maps or arrays, where x- and y-coordinates represent genotypes and environments, and the z-values represent phenotypic data using dots or other symbols and an appropriate color scheme to indicate the range of values. This paper illustrates such 'GenPhEn arrays' using data from a multi-location wheat trial (Triticum aestivum L.) and from simulations of response of common bean (Phaseolus vulgaris L.) to increased air temperature. By standardizing the values for various traits, the genotypic and environmental effects can be easily viewed and better comprehended, especially when presented in multi-trait arrays. The arrays allow presenting large amounts of data (i.e., 5000 data points or more) in a compact and readily interpretable fashion. Various modifications to the basic application are illustrated. It is expected that GenPhEn arrays can be used by plant breeders, agronomists, and others who are interested to rapidly examining large sets of data for patterns that merit further study using more quantitative approaches.