Neural networks trained on genetic text can 'read' the DNA of bacteria. Scientists took the powerful Evo 2 model and added a tiny, tunable sensor, a genetic probe, without touching the model itself. The probe picks up signals associated with dangerous genes, like antibiotic resistance. Similar to spectroscopy, a method of analyzing substances by their characteristic glow, the probe recognizes the model's specific 'response' to a genetic sequence.
This approach turned out to be surprisingly accurate. Even a simple probe gives high results, and with the addition of an attention mechanism, accuracy reaches almost 0.99—it can distinguish subtypes of drug resistance. And, like a metal detector, it handles fragments of DNA. This allows cheap and fast testing of hospital wastewater and hidden sewer threats, warning of infection outbreaks before symptoms appear in people.
as written by the authors arXiv:2607.14070
Genomic foundation models such as Evo 2 learn rich sequence representations, but their value for biosecurity screening is largely unexplored. We ask how much biosecurity-relevant signal is linearly accessible in these representations by training minimal linear and attention probes on frozen Evo 2 layer-26 activations, without fine-tuning the underlying model. Across held-out metagenomic test sets, the probes detect antimicrobial resistance (AMR) with strong discrimination: a linear probe reaches a region-level ROC-AUC of 0.888 (mean-pool), rising to 0.977 with a single-head attention probe. The probes resolve finer-grained AMR drug-class subcategories and separate them from unrelated functional genes, providing additional evidence that the learned signal is not explained solely by generic functional-gene status. Bacterial virulence is also decodable, though more weakly (region-level ROC-AUC 0.833). The AMR probe retains comparable ranking performance on simulated short reads without retraining, enabling evaluation before assembly in settings where assembly is computationally costly or unreliable. It achieves a read-level ROC-AUC of 0.898 (mean-pool), comparable to the mean-pooled full-region result. Within SynGenome, AMR-associated prompt labels are only weakly recoverable from Evo 1.5-generated sequences; these prompt-derived labels do not establish the function of the generated response sequences. A complementary sparse-autoencoder analysis recovers interpretable resistance-associated features but proves less consistent than the supervised probes. Together, these results position lightweight embedding-based probes as a fast, inexpensive first-pass detection layer for metagenomic biosurveillance and map both strengths and current limits of the approach. This work was conducted as part of the AIxBio Hackathon 2026 hosted by BlueDot Impact, Apart Research, and Cambridge Biosecurity Hub.
Antibiotic resistance is one of the major health threats: by 2050, it could kill 10 million people annually.