METAGENE-1
verifiedFreemetagene.ai
Open 7B-parameter metagenomic foundation model trained on wastewater DNA/RNA for pathogen detection and pandemic biosurveillance.
What it does
METAGENE-1 is a 7-billion-parameter autoregressive foundation model trained on over 1.5 trillion base pairs of DNA and RNA from human wastewater. Built by researchers at USC, Prime Intellect, and the Nucleic Acid Observatory, it targets pathogen detection and pandemic monitoring, and is released openly on GitHub and Hugging Face.
Core features
7B-parameter metagenomic language model
Trained on 1.5T+ base pairs from wastewater
Byte-pair tokenization tuned for genomic sequences
State-of-the-art pathogen-detection benchmarks
Open weights on Hugging Face and GitHub
Best for
→Pathogen and anomaly detection
→Biosurveillance and pandemic monitoring
→Metagenomic embedding and research
→Building scientific foundation models
Reviews
Big-picture takes: what it's for and whether it delivers. High-engagement YouTube videos — not sponsored.