aiqbench analysis
This paper introduces VitisGP, a curated benchmark that pairs genome-wide SNP data with 26 grapevine phenotype prediction tasks and compares classical models (logistic regression, Random Forest, XGBoost, CatBoost, GBLUP) with Transformers. CatBoost performs best overall, with a macro-average AUROC of 0.696, and sensory and colour traits are more predictable than agronomic ones. For ML readers, the main value is a harmonized benchmark for genotype-to-phenotype prediction on small, heterogeneous genomic data, where gradient-boosted trees still lead.
VitisGP: Multi-Task Benchmark for Genotype to Phenotype Grapevine Prediction
arXiv:2610.40668v1 Announce Type: cross Abstract: As climate change reshapes grape-growing conditions, predicting grapevine phenotypes from genomic variation is increasingly important for identifying and breeding cultivars suited to emerging environmental pressures. Existing research on the grapevine genotype-phenotype relationship using machine learning has often focused on individual traits, isolated regions, or a limited set of predictive models. We introduce VitisGP, a multi-phenotype curated dataset, integrating genome-wide SNP data with 26 heterogeneous phenotype prediction tasks spanning sommelier-curated aroma traits (N=11), botanical traits (N=14) and wine color---across 1,197 Vitis vinifera cultivars. We employ VitisGP to benchmark phenotype prediction, comparing machine-learning models with differing inductive biases---L2-regularized logistic regression, Random Forest, XGBoost, CatBoost, GBLUP and Transformer-based models, providing the first systematic comparison of this breadth on harmonized grapevine data. Across the benchmark, CatBoost achieves the strongest overall performance (macro-average AUROC=0.696 across 26 tasks), although model performance varies substantially across phenotype classes, with sensory and color traits showing stronger predictability than agronomic characteristics. VitisGP provides the community with a resource for investigating the genomic basis of phenotypic traits and screening grapevine diversity for climate-relevant characteristics.
Abstract:As climate change reshapes grape-growing conditions, predicting grapevine phenotypes from genomic variation is increasingly important for identifying and breeding cultivars suited to emerging environmental pressures. Existing research on the grapevine genotype-phenotype relationship using machine learning has often focused on individual traits, isolated regions, or a limited set of predictive models. We introduce VitisGP, a multi-phenotype curated dataset, integrating genome-wide SNP data with 26 heterogeneous phenotype prediction tasks spanning sommelier-curated aroma traits (N=11), botanical traits (N=14) and wine color---across 1,197 Vitis vinifera cultivars. We employ VitisGP to benchmark phenotype prediction, comparing machine-learning models with differing inductive biases---L2-regularized logistic regression, Random Forest, XGBoost, CatBoost, GBLUP and Transformer-based models, providing the first systematic comparison of this breadth on harmonized grapevine data. Across the benchmark, CatBoost achieves the strongest overall performance (macro-average AUROC=0.696 across 26 tasks), although model performance varies substantially across phenotype classes, with sensory and color traits showing stronger predictability than agronomic characteristics. VitisGP provides the community with a resource for investigating the genomic basis of phenotypic traits and screening grapevine diversity for climate-relevant characteristics.
Genomics (q-bio.GN); Machine Learning (cs.LG)
arXiv:2610.40668 [q-bio.GN]
(or arXiv:2610.40668v1 [q-bio.GN] for this version)
https://doi.org/10.48550/arXiv.2610.40668 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
40th Conference on Neural Information Processing Systems (NeurIPS 2026). Workshop: Tackling Climate Change with Machine Learning
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 285 words
Related coverage
Structured Multitask Gaussian Processes for Probabilistic Full-Body Human Motion Prediction
arXiv · Sep 28, 2026
DPTM-DT: Dual-Pretrained Transformer Multitask Representation Learning for Drug-Target Prediction
arXiv · Sep 22, 2026
FusionMMT: A Unified Multimodal and Multitask Learning Framework for Nuclear Fusion
arXiv · Sep 24, 2026
GRADE: Generalizable Reasoning-Aware Dialogue Evaluation for AI Tutors
arXiv · Jun 5, 2026