What happened
Microbial genomes encode compact molecular machines and diverse non-coding RNAs (ncRNAs) essential to gene regulation, pathogenesis, and many foundational biotechnologies. However, annotation remains largely protein-centric and homology-driven. Here, we introduce Minerva, a framework for coevolutionary mining that uses genome language models to predict local interactions directly from sequence as two-dimensional maps. Introducing two complementary techniques, categorical Jacobian fingerprinting and interaction heads, we demonstrate fast, accurate, alignment-free prediction of ncRNA base-pairing, monomeric protein contacts, and repetitive sequence motifs. Applied to 150 bacterial genomes, Minerva recovers known systems and predicts that 84.3% of predicted intergenic base-pairing falls outside of known annotations. In Pseudomonas, we find that the widespread TwoAYGGAY ncRNA family carries large secondary-structure extensions and is often flanked by short upstream repetitive motifs and larger downstream genomic repeats. Interpreting the coevolution maps, we find that Minerva emergently detects open-reading-frame (ORF) signatures at the DNA level despite never being trained to do so. In prophages within these genomes, we discover that Unknown Group 27 (UG27) reverse transcriptase systems encode arrays of structurally conserved yet sequence-diverse ncRNAs that template complementary DNA (cDNA) hairpin products. Together, these results establish coevolutionary mining as a scalable route to genome annotation and biological discovery across the rapidly expanding microbial universe. Update (2026-09-24): The phage stress test: a proving ground for genome language models in biological prediction . Reliable biological prediction by AI is most likely to emerge first in the simplest systems with rich datasets and the fastest opportunities for testing and refinement. Phages with small genomes are an ideal testbed. Decades of experimental work have created an unusually information-rich literature benchmark that sits largely outside the sequence repositories typically used to train genome language models (GLMs). Recent work has shown that GLMs such as Evo2 can generate viable whole bacteriophage genomes, demonstrating that genome-scale biological design is possible. The next question is more mechanistic: can such models correctly predict the effects of simple, local sequence changes? Here, we evaluated Evo2 against experimental mutation data from two model phages: the single-stranded RNA phage MS2 and the single-stranded DNA phage {Phi}X174. We first addressed whether Evo2 sequence log-likelihood scores could distinguish viable from nonviable mutations, and then whether models trained on Evo2 embeddings were predictive of function. Across both phages, Evo2 captu Update (2026-09-24): EPIC: An open community challenge for sequence-based prediction of transcription initiation in five non-model metazoans . Predicting gene expression from DNA sequence is a central problem with critical biological and clinical implications. Recent sequence-to-function models are reported to achieve improved performance, yet it remains unclear how much of what they learn reflects genuine principles of eukaryotic transcription initiation and to what extent they are able to generalize beyond humans and other primary model species. A fair and blind benchmark has likewise been missing. Here we introduce EPIC, the Eukaryotic Promoter and transcription Initiation prediction Challenge. Teams receive strand-specific, single-nucleotide-resolution initiation profiles for 80-95% of the genome and are asked to predict the remainder from DNA sequence alone. To level the field, the challenge relies on understudied animals spanning three phyla: octopus, oyster, milkweed bug, Indian meal moth, and shark. EPIC is open to everyone and closes on December 31, 2026. All teams clearing the dinucleotide precision baseline are invited to join the consortium authorship of the post-