Evidence Brief 2024-01-10
AlphaFold and the protein folding revolution
How deep learning solved a 50-year-old grand challenge in biology.
Read article → Flagship Hub
AI × Biology
AI is changing biology by turning messy, high-dimensional data into models that can predict, compress, and generate hypotheses—from molecules to cells to brains. This hub tracks what works, what's uncertain, and what the evidence actually supports.
Where algorithms meet organisms
Predict
Models that predict molecular structures, cellular states, and clinical outcomes.
Compress
Learning efficient representations of complex biological systems.
Generate
AI systems that design proteins, suggest experiments, and form hypotheses.
Hub Sections
Data Foundations
- Genomics and sequencing data
- Single-cell datasets
- Medical imaging and clinical records
- Omics integration challenges
- Data governance and privacy
Molecules & Proteins
- Structure prediction (AlphaFold)
- Protein design and engineering
- Drug discovery pipelines
- Molecular dynamics simulations
- Ligand binding and docking
Systems Biology
- Gene regulatory networks
- Causal inference methods
- Multiscale modeling
- Cell signaling pathways
- Population and evolutionary dynamics
Neuroscience & Cognition
- Brain data analysis
- Neural network models
- Brain-computer interfaces
- AI-inspired neuroscience
- Caveats on brain-AI analogies
Translational Medicine
- Diagnostic tools and triage
- Clinical decision support
- Trial design and stratification
- Real-world evidence
- Deployment challenges
AI Safety & Biosecurity
- Model evaluation standards
- Misuse risk assessment
- Dual-use considerations
- Safeguards and oversight
- Policy interfaces
Content Formats
Each format serves a different purpose in navigating this complex intersection.
Model Cards
What a model was trained on, what it predicts, and its known failure modes.
Dataset Spotlights
Structure, biases, access constraints, and reuse tips for key datasets.
Paper-to-Practice
What changed since last year, and what still fails in real-world deployment.
Benchmarks
Does the result hold across labs, datasets, and real-world conditions?
Recent Articles
View All Methods 2024-01-08
What model cards should look like in biology
Documenting training data, intended use, and known limitations.
Read article → Explainer 2024-01-05
Single-cell RNA-seq: a practical guide
Understanding preprocessing, integration, and cell type annotation.
Read article → Analysis 2024-01-02
The limits of AI in drug discovery
Where generative models succeed, struggle, and what comes next.
Read article →Evidence and Uncertainty
AI in biology moves fast. We label what is well-supported by evidence versus what is promising but unproven. We document training data, evaluation benchmarks, and known failure modes. We distinguish hype from what's actually deployed.
Explore the other flagship hub
If you're interested in AI and biology, you may also find our Quantum ↔ Cosmic hub compelling—a deep dive into the physics of reality across scales.
Visit Quantum ↔ Cosmic