Wine fermentation remains inherently variable because of the genetic and phenotypic diversity of Saccharomyces cerevisiae and non- Saccharomyces yeasts, microbial interactions, and climate-driven shifts in grape composition, challenging predictable and low-intervention winemaking. Recent population genomics advances, including telomere-to-telomere assemblies and pan-genome analyses of yeast genomes, have transformed the field by revealing structural variation, introgressions, hybridization, and gene content diversity underlying key enological traits. High-throughput functional genomics, quantitative trait locus mapping, multi-omics, and machine learning increasingly connect these features with fermentation kinetics, stress tolerance, and aroma biosynthesis. The near completion of the Synthetic Yeast Genome (Sc2.0) and its SCRaMbLE system further expands the experimental design space for rapid genome rearrangement and strain innovation. These advances have improved identification of candidate determinants of industrially relevant phenotypes, but robust genotype-to-phenotype prediction remains limited by polygenic architectures, epistasis, environmental dependence, and microbial context. Future progress will depend on integrating population genomics with functional validation, realistic phenotyping, and interpretable predictive frameworks to support rational yeast engineering and more consistent, sustainable winemaking.
Kutyna et al. (Sat,) studied this question.