Torsten Pook

The paper develops a maize MAGIC population to dissect grain yield and proxy traits under stress, identifying QTL and improving genomic prediction models.

This paper optimizes wheat breeding programs using an evolutionary algorithm, achieving higher genetic gain and diversity while maintaining a fixed budget.

This paper quantifies heat stress impact on dairy cattle using a novel pipeline and identifies causal genomic regions associated with heat tolerance traits.

The paper uses stochastic simulations to optimize a maize breeding scheme by balancing genetic gains and diversity through additional phenotyping and external genetic material.

The paper compares selection strategies to improve genetic gain and reduce inbreeding over multiple generations using simulation software.

The paper develops optimized matrix-vector multiplication methods for genomic prediction, improving computational efficiency on GPUs and CPUs for large datasets.

This paper introduces a framework for numerically optimizing breeding programs by maximizing a target function that combines genetic gain and diversity conservation, using stochastic simulations and kernel regression.

This paper evaluates imputation methods for low-density marker data in sugar beet breeding, finding that AlphaPlantImpute2 performs well with default settings while Beagle benefits from parameter tuning, and combining both tools yields the highest accuracy.

The paper develops an eggplant MAGIC population to identify genes associated with anthocyanin pigmentation, finding strong associations with MYB113 and COP1 genes.

The paper introduces HBimpute, an imputation pipeline that improves calling accuracy, coverage, and read depth from low-read-depth sequencing data, achieving results comparable to high-density arrays.

The paper compares breeding strategies for creating a synthetic pig line, finding that combining genomic diversity management with selection starting at G4 and moderate diversity control preserves diversity and achieves genetic progress.

The paper compares genomic prediction models for maize traits, finding that selecting top SNP interactions improves prediction accuracy compared to additive models.

The paper investigates how SNP array design affects ascertainment bias by redesigning a chicken SNP array and analyzing allele frequency spectra and heterozygosity estimates.

MoBPSweb simulates and compares breeding programs by integrating a unified language with simulation software, evaluating scenarios based on genomic gains, inbreeding rates, and accuracy of breeding value estimation.

This paper compares bivariate genomic prediction models for plant traits across two growing seasons, finding that models incorporating epistasis improve prediction accuracy.

This paper proposes local convolutional neural networks for genomic prediction, showing improved performance over traditional methods in most cases, especially with medium complexity traits and large datasets.

This paper proposes a general definition of animal breeding programs using a modular graph representation and introduces a web tool and R-package for simulating and analyzing breeding programs.

MoBPS simulates breeding programs and compares their economic and genetic impacts using a modular, efficient framework with customizable modules.

The paper introduces two methods for analysis of variance in high-dimensional data and evaluates their performance on genomic datasets.

The paper evaluates BEAGLE versions for imputation quality in crop and livestock data, finding that parameter tuning significantly improves imputation accuracy.

HaploBlocker identifies subgroup-specific haplotype blocks and constructs libraries to represent genetic variability with fewer parameters than traditional methods.