Replication Data for: Nonlinear Genomic Selection Index Accelerates Multi-Trait Crop Improvement

The data in this dataset are used to explore the Quadratic Genomic Selection Index (QGSI), a genomic extension of the quadratic phenotypic selection index (QPSI) that integrates genomic estimated breeding values (GEBVs) within a unified quadratic framework. Classical linear phenotypic (LPSI) and genomic (LGSI) selection indices assume additivity and linearity. This can limit their ability to exploit nonlinear trait relationships. QGSI combines additive, squared, and cross-product terms of GEBVs. It enables phenotype-free, rapid-cycle multi-trait selection while capturing genome-wide nonlinear relationships. Two genomic prediction strategies were used to evaluate QGSI: (i) a maximum-likelihood additive genomic model and (ii) a nonlinear multi-trait Gaussian kernel (reproducing kernel Hilbert space, RKHS) model that accommodates epistatic signals. Analyses were performed using multiple real wheat datasets. The results of the comparisons are reported in the accompanying article.