{"id":26098036,"url":"https://github.com/deepbiolab/neural-bio","last_synced_at":"2026-04-18T22:32:02.618Z","repository":{"id":280144925,"uuid":"941099994","full_name":"deepbiolab/neural-bio","owner":"deepbiolab","description":"A hybrid modeling framework combining neural networks with physics-based constraints for bioreactor process optimization and control.","archived":false,"fork":false,"pushed_at":"2025-03-01T14:30:53.000Z","size":10435,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-03-01T15:26:58.769Z","etag":null,"topics":["bioreactor","biotechnology","hybrid-modeling","neural-networks","physics-informed-neural-networks","pinn","process-optimization"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/deepbiolab.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2025-03-01T13:42:39.000Z","updated_at":"2025-03-01T14:32:25.000Z","dependencies_parsed_at":"2025-03-01T15:37:36.807Z","dependency_job_id":null,"html_url":"https://github.com/deepbiolab/neural-bio","commit_stats":null,"previous_names":["deepbiolab/neural-bio"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepbiolab%2Fneural-bio","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepbiolab%2Fneural-bio/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepbiolab%2Fneural-bio/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepbiolab%2Fneural-bio/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/deepbiolab","download_url":"https://codeload.github.com/deepbiolab/neural-bio/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":242714052,"owners_count":20173581,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["bioreactor","biotechnology","hybrid-modeling","neural-networks","physics-informed-neural-networks","pinn","process-optimization"],"created_at":"2025-03-09T15:59:41.284Z","updated_at":"2026-04-18T22:31:57.574Z","avatar_url":"https://github.com/deepbiolab.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\n# Neural-Bio: Hybrid Modeling Framework for Bioreactor Optimization\n\n![](./assets/arch.svg)\n\nA hybrid modeling framework combining neural networks with physics-based constraints for bioreactor process optimization and control.\n- Neural network for reaction rate prediction\n- First principle model for bioreactor dynamics\n- Physics-informed constraints integration\n- Adaptive weighted loss function\n\n## Overview\n\n![](./logs/training_animation.gif)\n\n🧬 Neural-Bio is a PyTorch-based framework that implements a hybrid modeling approach for bioprocess optimization. It combines deep learning with mechanistic models to predict and optimize bioreactor processes, incorporating both data-driven insights and physical constraints.\n\n\n## Installation\n\n```bash\n# Clone the repository\ngit clone https://github.com/deepbiolab/neural-bio.git\ncd neural-bio\n\n# Create a virtual environment (optional but recommended)\npython -m venv venv\nsource venv/bin/activate  # On Windows: venv\\Scripts\\activate\n\n# Install dependencies\npip install -r requirements.txt\n```\n\n## Usage\n\n### Training\n\n```bash\npython main.py --mode train\n```\n\n- Place training data files in the `dataset/train` folder\n- Ensure data follows the required format (see Data Format section)\n- Adjust configuration settings in `src/config/default_config.py` as needed\n- Training and validation performance imgs are saved in the `logs/` directory\n\n### Testing\n\n```bash\npython main.py --mode test\n```\n\n- Place test data files in the `dataset/test` folder\n- Ensure data follows the required format (see Data Format section)\n- Verify model checkpoint exists in `checkpoints/` directory\n- Test performance is saved in the `logs/` directory\n\n### Prediction\n\n```bash\npython main.py --mode predict\n```\n\n- Place input data files in the `dataset/predict` folder\n- Ensure data follows the required format (see Data Format section)\n- Verify model checkpoint exists in `checkpoints/` directory\n- Prediction results are saved in the `results/` directory\n\n## Project Structure\n\n```\nneural-bio/\n├── src/\n│   ├── config/          # Configuration settings\n│   ├── data/            # Data processing and dataset classes\n│   ├── models/          # Neural network and hybrid model implementations\n│   ├── trainer          # Training utilities\n│   ├── inference        # inference utilities\n│   └── utils/           # Visualization and evaluation tools\n├── dataset/             # Data directory (not included)\n├── checkpoints/         # Model checkpoints\n├── results/             # Prediction results\n├── logs/                # Training logs\n└── main.py              # Main script for training, prediction, and testing\n```\n\n## Configuration\n\nThe model behavior can be customized through the configuration file `src/config/default_config.py`. Key parameters include:\n\n- Dataset parameters (time steps, variables, etc.)\n- Model architecture (layers, dimensions)\n- Training parameters (learning rate, batch size)\n- Loss function weights\n- Evaluation metrics\n\n## Data Format\n\nThe framework expects two main data files in CSV format:\n\n### OWU (Observation Wise Unit) Data\nTime-series measurements of process variables for each experimental run.\n\n```\nrun,time,VCD,Glc,Lac,Titer,Glc_feed\n```\n\n- `run`: Experiment run index\n- `time`: Time point (days)\n- State Variables:\n  - `VCD`: Viable Cell Density (10⁶ cells/mL)\n  - `Glc`: Glucose concentration (g/L)\n  - `Lac`: Lactate concentration (g/L)\n  - `Titer`: Product titer (g/L)\n- Feeding Variables:\n  - `Glc_feed`: Glucose feeding rate (g/L/day)\n\n### DOE (Design of Experiments) Data\nExperimental design parameters and initial conditions for each run.\n\n```\nrun,feed_start,feed_end,Glc_feed_rate,Glc_0,VCD_0\n```\n\n- `run`: Experiment run index\n- Feeding Schedule:\n  - `feed_start`: Day to start feeding\n  - `feed_end`: Day to stop feeding\n  - `Glc_feed_rate`: Glucose feed concentration\n- Initial Conditions:\n  - `Glc_0`: Initial glucose concentration\n  - `VCD_0`: Initial viable cell density\n\n### Data Organization\n\n```\ndataset/\n├── interpolation/\n│   ├── train/\n│   │   ├── owu.csv\n│   │   └── owu_doe.csv\n│   ├── test/\n│   │   ├── owu.csv\n│   │   └── owu_doe.csv\n│   └── predict/\n│       ├── owu.csv\n│       └── owu_doe.csv\n```\n\n## Contributing\n\nContributions are welcome! Please feel free to submit a Pull Request.\n\n## License\n\nThis project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.\n\n## Citation\n\nIf you use this code in your research, please cite:\n\n```bibtex\n@software{neural_bio2025,\n  author = {Tim-Lin},\n  title = {Neural-Bio: Hybrid Modeling Framework for Bioreactor Optimization},\n  year = {2025},\n  publisher = {GitHub},\n  url = {https://github.com/deepbiolab/neural-bio}\n}\n```\n\n## Contact\n\nFor questions and feedback, please open an issue on GitHub.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdeepbiolab%2Fneural-bio","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdeepbiolab%2Fneural-bio","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdeepbiolab%2Fneural-bio/lists"}