Environment ensemble models for genomic prediction in common bean ( <i>Phaseolus vulgaris</i> L.)
Bibliographic record
Abstract
For important food crops such as the common bean (Phaseolus vulgaris, L.), global demand continues to outpace the rate of genetic gain for quantitative traits. In this study, we leveraged the multi-environment trial (MET) dataset from the cooperative dry bean nursery (CDBN) to investigate the use of ensemble models for genomic prediction. This set spans 70 locations and 30 years, and accounts for over 150 phenotypes and hundreds of genotypes sequenced for 1.2 million single nucleotide polymorphism markers. We tested three models (linear regression, ridge regression, and neural networks). Each of the three models was implemented using three different approaches: (1) combining all data into one model (singular model), (2) all available single locations were used to train individual submodels comprising one ensemble model (ensemble model), and (3) optimized sets of single locations were used to train individual submodels comprising one ensemble model (optimized ensemble model). The optimized ensemble approach worked best for low-variance locations because the model variance was reduced by averaging across submodels in the ensemble. For models with low prediction accuracy, the ensemble approach can increase accuracy. In certain locations, prediction accuracy was able to overcome narrow-sense heritability, indicating that genomic selection is more efficient than phenotypic selection in these locations. This study indicates that breeding program collaboration can be a way to bypass the bottleneck of low data volume, as pooled data from the CDBN MET produced prediction accuracies of 0.70 for days to flowering, 0.54 for days to maturity, 0.95 for seed weight, and 0.67 for seed yield in individual locations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".