Phantom epistasis in genomic selection : on the predictive ability of epistatic models

• Autores: Schrauf, Matías Florián; Martini, Johannes W. R.; Simianer, Henner; Campos, Gustavo de los; Cantet, Rodolfo Juan Carlos; Freudenthal, Jan; Korte, Arthur; Munilla Leguizamón, Sebastián
• Año de publicación: 2020
• Idioma: inglés
• Tipo de recurso: artículo
• Fuente: G3: Genes, Genomes, Genetics; Vol.10, no.9; 3137-3145; http://www.g3journal.org/
• Materia: EPISTASIS; ADDITIVE EFFECTS; GENOMICS; BREEDING; GENPRED; GENOMIC PREDICTION; SHARED DATA RESOURCES
• Institución: Universidad de Buenos Aires. Facultad de Agronomía

Fil: Schrauf, Matías Florián. Universidad de Buenos Aires. Facultad de Agronomía. Buenos Aires, Argentina.
Fil: Martini, Johannes W. R. International Maize and Wheat Improvement Center (CIMMYT), Texcoco, Mexico.
Fil: Simianer, Henner. University of Göttingen. Department of Animal Sciences. Center for Integrated Breeding Research. Germany.
Fil: Campos, Gustavo de los. Michigan State University. Department of Epidemiology and Biostatistics. East Lansing, Michigan, EEUU.
Fil: Cantet, Rodolfo Juan Carlos. Universidad de Buenos Aires. Facultad de Agronomía. Buenos Aires, Argentina. - CONICET. Buenos Aires, Argentina.
Fil: Freudenthal, Jan. University of Würzburg. Center for Computational and Theoretical Biology (CCTB). Germany.
Fil: Korte, Arthur. University of Würzburg. Center for Computational and Theoretical Biology (CCTB). Germany.
Fil: Munilla Leguizamón, Sebastián. Universidad de Buenos Aires. Facultad de Agronomía. Buenos Aires, Argentina. - CONICET. Buenos Aires, Argentina.
Genomic selection uses whole-genome marker models to predict phenotypes or genetic values for complex traits. Some of these models fit interaction terms between markers, and are therefore called epistatic. The biological interpretation of the corresponding fitted effects is not straightforward and there is the threat of overinterpreting their functional meaning. Here we show that the predictive ability of epistatic models relative to additive models can change with the density of the marker panel. In more detail, we show that for publicly available Arabidopsis and rice datasets, an initial superiority of epistatic models over additive models, which can be observed at a lower marker density, vanishes when the number of markers increases. We relate these observations to earlier results reported in the context of association studies which showed that detecting statistical epistatic effects may not only be related to interactions in the underlying genetic architecture, but also to incomplete linkage disequilibrium at low marker density (“Phantom Epistasis”). Finally, we illustrate in a simulation study that due to phantom epistasis, epistatic models may also predict the genetic value of an underlying purely additive genetic architecture better than additive models, when the marker density is low. Our observations can encourage the use of genomic epistatic models with low density panels, and discourage their biological over-interpretation.
grafs.