How to Build an Open-Source Machine Learning Platform in Biology?
summary
OmicLearn is an open source web platform that runs tabular omics workflows with pandas/numpy, scikit‑learn and XGBoost, deployed via GitHub Actions and AWS with PyInstaller desktop builds and pytest integration tests. It saves session history export, produces publication‑quality Plotly figures for direct download, and already helps users rediscover known biomarkers (Tau in Alzheimer’s). The project hits reproducibility goals, earned “app of the month,” published a paper with ~4,000 views in six months and plans Optuna hyperparameter tuning and continued maintenance.
AI-written summary · about
chapters (8)
- 0:00Introduction and speaker background
- 1:29Basics of biology for ML
- 7:27Machine learning in biology
- 11:44Why build omiclearn
- 16:10Tool selection and architecture
- 20:08Project highlights and impact
- 22:00Demo of omiclearn platform
- 25:13Q&A session
related talks
01h 16mReproducible machine learning and science with python
40mInstructLab: Applying Open Source Methods to Building and Trai...
37mPlatform for Complete Machine Learning Lifecycle
45mOpen Source for Open Science with Frictionless Data
26mOracle Machine Learning: what's new and what's next?
16mKeynote: Accelerating AI/ML Development with Open Source Software
59mAlgorithmia in Action: Take your ML models from training to production
18mBuilding a Machine Learning Platform at