biofilmevolutionanalyzer - Biofilm Evolution Analyzer v1.0.1 – Compatibility & Setup Fixes

What’s new in v1.0.1

Bug-fix and quality-of-life update to make local installation smoother, especially on Python 3.13 / Apple-silicon.

✨ Highlights

1. TensorFlow now optional
   • utils/ml_models.py wraps the TF import in a try/except, creates safe placeholders, and raises an informative error only when the user selects “Neural Network”.
   • All other ML algorithms run without TensorFlow.

2. Updated dependencies
   • requirements.txt – commented out the TensorFlow line (no wheels for Py 3.13); everything else installs cleanly with pip install -r requirements.txt.

3. Expanded Setup Guide
   • New “Running in a Virtual Environment” section explains PEP 668, Homebrew Python, and TF work-arounds.
   • Troubleshooting tips for ModuleNotFoundError and Apple-silicon notes.

🛠 Internal changes

• Added TENSORFLOW_AVAILABLE flag and placeholder class for graceful degradation.
• Guarded neural-network builder functions to error early when TF is missing.
• Committed and pushed tag v1.0.1.

🚀 Upgrade instructions

```bash git pull git checkout v1.0.1 # or just stay on main (same commit)

python3 -m venv .venv source .venv/bin/activate pip install -r requirements.txt streamlit run app.py ``` Select any algorithm except “Neural Network” if TensorFlow isn’t installed.

🔗 Citation

The Zenodo record for v1.0.1 will appear shortly (auto-sync). Cite using the updated DOI once published.

---

Thanks to everyone testing the early release! Please open issues for any remaining setup hurdles.

- Python
Published by mojo8787 8 months ago

biofilmevolutionanalyzer - Biofilm Evolution Analyzer v1.0.0

Overview

Biofilm Evolution Analyzer is an open-source Streamlit platform for integrative, multi-omics analysis of bacterial lifestyle transitions.
Version v1.0.0 is the first publicly archived release and now citable via DOI (generated by Zenodo on publication).

Key Features

• Multi-omics data import – transcriptomics, Tn-Seq, and phenotype CSVs
  
• Interactive exploration – summary stats, PCA, heatmaps, correlation plots
  
• Machine-learning module – classical ML and neural-network models with feature importance analysis
  
• Metabolic modeling – integrates COBRA genome-scale models, flux balance analysis, trade-off exploration
  
• In-silico evolution – simulate adaptive trajectories under custom environmental constraints
  
• Automated experiment design – Bayesian optimisation and factorial/RSM designs
  
• Reusable utilities – modular functions in utils/ for data processing, modelling, visualisation
  

What’s new in v1.0.0

• Added CITATION.cff with complete metadata (ORCID [0000-0003-2070-2811

- Python
Published by mojo8787 8 months ago