Building the Future of Biological Knowledge
Flash-P is more than a tool — it's a platform where AI agents and domain experts co-create validated biological knowledge. Here is where we are and where we're going.
Desktop App & AI Agents
Flash-P is available as a desktop application. Five specialized AI agents work together to mine literature, construct causal networks, and validate them — all from a single query on your machine.
Desktop Application
Full pipeline running locally. No cloud, no cluster. Just your laptop, Python, and an internet connection for AI features.
- Windows, macOS, Linux
- Auto-updates via Tauri
- Offline mode with cached data
- Claude Code integration
5 Specialized Agents
Curator, Perturbation, Builder, Validator, Refinement — each expert in a stage of the pipeline.
- Literature mining with DOIs
- Automated perturbation datasets
- Network construction with equations
- Iterative improvement
3 Validation Methods
Every network is validated using algebraic rules, ODE with Hill functions, and Random Walk with Restart.
- Flash-P algebraic propagation
- ODE with Hill functions
- Random Walk with Restart
- Concordance scoring
Community Hub & Shared Validation
An open platform where labs share validated networks, contribute perturbation data, and review edges. Every validation makes every network better for everyone.
Network Sharing & Discovery
Publish validated networks to the Hub. Browse networks by species, trait, and accuracy. Download complete networks with all edges, equations, and validation results.
- ORCID-linked contributions
- Citation tracking
- Leaderboard & recognition
- Community-driven quality control
Perturbation Data Pooling
Every perturbation experiment shared to the Hub strengthens the validation pool. Flash-P agents automatically access community data to build better networks.
Tools, Plugins & Applications
An ecosystem of plugins that connect genomic data to biological mechanisms — making Flash-P networks actionable for breeders and translational researchers.
SNP-to-Node Mapping
Connect genomic variants (SNPs, QTLs) to network nodes via MCP plugins. Breeders can see which markers affect which regulatory pathways.
Breeder Applications
Translate network knowledge into breeding decisions. Simulate gene stacking, predict trait trade-offs, and identify optimal crossing targets.
- Gene stacking simulation
- Trait trade-off prediction
- Marker-assisted selection targets
- Cross-trait interaction warnings
Plugin Ecosystem
Connect Flash-P to external databases and tools through MCP and APIs. Each plugin adds a new data layer to your network.
The Evolving Agent Loop
Every contribution feeds back into the system. Flash-P agents don't just build networks — they learn from every correction, every validation, every expert comment.
Scientists build networks
Labs validate experimentally
Experts review edges
Community votes on quality
Agent learns from feedback
Better networks produced
Start Building Networks Today
Download Flash-P and start building validated causal networks. The community hub and tools are on the way — for now, the desktop app has everything you need.