Accessible
Run machine learning workflows through Galaxy forms and histories without requiring local command-line setup.
Galaxy Learning and Modeling
GLEAM
No-code machine learning tools for the Galaxy computational workbench, spanning tabular modeling, image learning, and multimodal prediction.
Built for reproducible biomedical ML
Train, evaluate, and share models in Galaxy.
Reproducible
Capture inputs, parameters, model artifacts, reports, metrics, and configuration files as Galaxy datasets.
Extensible
Use maintained wrappers and containerized runtimes that can be installed into public, institutional, or development Galaxy servers.
Workbench
GLEAM software tools
Structured data
Tabular Learner
Train and compare classification or regression models with PyCaret, generate evaluation reports, and reuse the best model for prediction.
Computer vision
Image Learner
Train image classification or regression models from image ZIP archives and metadata tables using Ludwig, TorchVision, and MetaFormer backbones.
Mixed modality
Multimodal Learner
Combine tabular, text, and image features in AutoGluon Multimodal with configurable splits, metrics, quality presets, and HTML reporting.
Galaxy Training Network
Hands-on GLEAM tutorials
Tabular Learner
Predict patient response to therapy using automated machine learning with clinical data
Train a tabular immunotherapy-response classifier, compare candidate models, and evaluate discrimination, calibration, and threshold behavior against published results.
- Published
- May 2025
- Level
- Intermediate
- Time
- 1 hour
Image Learner
Skin lesion classification using computer vision models
Prepare a balanced HAM10000 subset, run Image Learner with a pretrained CaFormer deep learning backbone, and evaluate accuracy, weighted precision, recall, F1, and confusion patterns.
- Published
- January 2026
- Level
- Intermediate
- Time
- 1 hour
Multimodal Learner
Head and neck cancer recurrence prediction using clinical and imaging data
Train a late-fusion model from clinical tabular fields, text columns, and CD3/CD8 image archives, then interpret ROC, PR, confusion matrix, calibration, and threshold-dependent metrics.
- Published
- March 2026
- Level
- Intermediate
- Time
- 1 hour