Guillaume P. Ramstein

This study uses temporal allele frequency changes to detect adaptive variants in winter wheat breeding programs, finding that selection signals identify non-neutral variants associated with grain yield and protein content.

This study uses ESM2 to identify beneficial genetic variants in a Sorghum bicolor diversity panel by analyzing PRC scores and their association with agronomic traits.

The paper validates variant effect predictions in a Brachypodium distachyon mutant population using burden and purging tests, showing ESM outperforms SIFT and PlantCAD for missense variants.

The paper develops deep learning models to predict gene expression across species and assesses their ability to capture regulatory variant effects in Brachypodium mutant lines, showing improved accuracy over existing methods.

The paper evaluates genomic prediction models for agronomic traits in perennial ryegrass under genotype x environment interactions, showing improved accuracy when accounting for these interactions.

The paper predicts evolutionary constraint in maize using genomic annotations to prioritize causal variants for fitness-related traits, improving genomic prediction accuracy.

The paper integrates GWAS and TWAS to identify 22 candidate genes regulating maize leaf cuticular conductance, with ISTL1 validated as a key gene through near-isogenic lines.

This paper introduces HARE, a method for imputing cis haplotype-associated RNA expression to improve genomic prediction of complex traits across maize populations.

This study shows dominance and gene annotations improve genomic prediction of agronomic traits in maize hybrids.

The paper develops deep learning methods that incorporate evolutionary relationships to predict transcript abundance from DNA sequence, achieving auROC values between 0.72 and 0.94.