CarveFungi
Automatic metabolic model reconstruction of fungi
Reconstruction
CarveFungi Metabolic Model Reconstruction - Reconstruct genome-scale metabolic models from genomic annotations
Features
- Genome annotation import
- Subcellular localization prediction import
- Gap-filling optimization
- SBML model export
Visualization
CarveFungi Metabolic Model Visualization - Interactive visualization of fungal metabolic models powered by CarveFungi
Features
- Interactive network visualization
- FBA optimization analysis
- Custom compound filtering
- Export to PNG, JSON & HTML
GPR Rules Maker
GPR Rules Fixing - Analyze and improve Gene-Protein-Reaction (GPR) rules using BLAST-based genetic context analysis
Features
- BLAST-based genetic context analysis
- GPR rule validation
- Complex assignment
- Rule optimization report
GECKO
GECKO Model Conversion - Add enzyme constraints to your genome-scale metabolic model using GECKO for improved phenotype predictions
Features
- Enzyme-constrained model generation
- Optional proteomics integration
- Protein pool constraints
- SBML model export
Constraint Editor
Modify reaction bounds on your metabolic model and run FBA simulations interactively
Features
- Edit reaction bounds interactively
- FBA simulation with custom objectives
- Download constrained model
- Search & filter reactions
About CarveFungi
Automated Reconstruction of Compartmentalized Fungal Metabolic Models
What is CarveFungi?
CarveFungi is an automated computational platform for rapid reconstruction of genome-scale, compartmentalized metabolic models (GEMs) for fungi directly from protein sequences. Unlike existing automated reconstruction tools, CarveFungi explicitly accounts for subcellular compartmentalization, a defining feature of eukaryotic metabolism that strongly affects metabolic connectivity, pathway feasibility, and phenotype prediction accuracy.
Key Technologies
- Deep Learning: Protein subcellular localization prediction
- Model Carving: Top-down approach from universal fungal network
- Compartmentalization: Explicit subcellular localization modeling
- High-Throughput: Species-specific model generation at scale
Impact & Results
- 800+ Fungal Species: Comprehensive coverage of fungal kingdom
- Biologically Meaningful: Models reflect phylogenetic relationships
- Metabolic Diversity: Captures species-specific pathways
- Eukaryotic Features: Accounts for compartmentalized metabolism
- Accurate Predictions: Improved phenotype prediction accuracy
How It Works
- Predicts subcellular localization of proteins (This web server doesn't do the compartmentalization prediction). Please, use this standalone tool for compartmentalization prediction.
- Scores proteins based on protein function and localization. (Annotation can be done using sequence based tools like EggNog or structural based tools like ActSeekN)
- Maps genes to metabolic reactions
- Integrates with universal fungal metabolic network
- Carves species-specific models
- Generates SBML-formatted GEM
Applications
CarveFungi models enable:
- Metabolic engineering in fungi
- Pathway optimization studies
- Phenotypic predictions
- Industrial strain improvement
- Systems biology analysis
- Comparative metabolomics
This algorithm has been developed in collaboration with Paula Jouhten
This work has been partially funded by the NSF Food ID project
CarveFungi • Automatic Metabolic Model Reconstruction Platform
Developed at VTT Technical Research Centre of Finland