The research, led by Avery J. C. Noonan and colleagues at Lawrence Berkeley National Laboratory and collaborating institutions, was built on an unusually rigorous foundation. The team systematically optimized their workflow over more than 13.2 million individual training runs, evaluating over 200 parameter settings across six published phage–host interaction datasets. Together these datasets encompass 115,037 experimentally determined interactions spanning 949 bacterial strains and 518 phages, drawn from Escherichia coli, Klebsiella, Pseudomonas and Vibrionaceae hosts. The proportion of infectious interactions within each dataset ranged from as little as 2.1 percent to 36.2 percent, a severe imbalance that mirrors the reality of phage biology, where most phages fail to infect most strains they encounter.

The core technical challenge the researchers faced is one familiar to anyone working at the intersection of genomics and machine learning: dimensionality. Each bacterial or phage genome can contribute thousands of numerical features—protein families, gene fragments or sequence k-mers—while the phenotypic data describing which phage infects which strain is sparse, binary and heavily skewed toward negative outcomes. This combination invites overfitting, in which a model memorizes quirks of its training data rather than learning generalizable rules. Many earlier tools sidestepped the problem by restricting features to genes already known to mediate infection, such as phage tail fibres or host capsule-synthesis loci. That strategy works within a single, well-studied genus but cannot transfer to emerging pathogens or poorly characterized phages, and it forecloses the discovery of novel infection determinants.

The new framework, by contrast, is deliberately phylogeny-agnostic. Genomes are represented as binary presence–absence profiles of protein families, clustered with MMSeqs2 at optimized similarity thresholds. Informative features are then selected through recursive feature elimination, and predictions are generated by an ensemble of gradient-boosted decision trees using the CatBoost algorithm, with class weights computed separately for each phage to compensate for imbalance. The team compared eight machine-learning algorithms, six feature-selection methods and multiple genome-representation schemes, finding that algorithmic choices mattered more than how genomes were encoded. Notably, the simplest representation—short 3-amino-acid k-mers—performed worst, while protein families, longer k-mers and hybrid schemes captured sufficient biological signal.

Performance was assessed through nested cross-validation under three increasingly demanding scenarios: predicting infections of unseen bacterial strains by known phages, infections of known strains by unseen phages, and interactions between entirely novel phage–strain pairs. Across datasets, the area under the receiver operating characteristic curve ranged from 0.67 to 0.94, with Matthews correlation coefficients between 0.13 and 0.54. For E. coli, the model achieved an AUROC of 0.87, essentially matching the 0.86 reported by a previous species-specific method that required phage-specific models and host features based on known E. coli infection mediators. The new approach reached comparable accuracy without any of those constraints. Performance correlated strongly with dataset size, and the weakest results came from the smallest, most imbalanced dataset—a Klebsiella matrix of only 3,658 interactions with 4.9 percent positives—suggesting the limits reflect data availability rather than flaws in the method.

The framework also translated its probability scores into practical phage-cocktail design. Phages were clustered into ‘activity groups’ using HDBSCAN density-based clustering, and the top-ranked phage from each cluster was selected, balancing predicted efficacy against mechanistic diversity. In silico tests across twenty-fold cross-validation showed that model-guided selection outperformed the common heuristic of choosing the most promiscuous phages. For single-phage selection, the models identified active phages for 66.9 percent of E. coli strains, 66.7 percent of Klebsiella strains and 39.4 percent of Vibrionaceae strains—improvements of 1.1-fold, 3.1-fold and 2.7-fold over promiscuity-based choices. Five-phage cocktails achieved bacterial coverage of up to 97.5 percent for E. coli, 87.8 percent for Klebsiella and 57.5 percent for Vibrionaceae, consistently beating baseline strategies.

Crucially, the team did not stop at computational validation. Because the phages in their training data were not available for retesting, they experimentally challenged the model with 52 previously unseen phages from the BASEL collection, spotted against 25 E. coli strains from the ECOR reference collection. Of the 1,300 tested interactions, 1,240 yielded clear phenotypes, and the model’s predictions against this genuinely novel matrix achieved an AUROC of 0.84—only marginally below its cross-validation performance of 0.87. This demonstrated that the framework generalizes to phages it had never encountered and to interaction data generated in a different laboratory with different protocols, a stringent test that many published prediction tools have never undergone.

To verify that the model’s predictions rest on real biology rather than statistical artifacts, the researchers turned to genome-wide random barcode transposon-site sequencing, or RB-TnSeq. They screened a pooled library of 3,804 gene knockouts in E. coli ECOR27 against 19 phages, identifying 51 genes whose disruption altered susceptibility. When these genetically validated infection mediators were compared with the model’s top predictive features, ranked by SHAP values, 68.6 percent showed a connection: 12 mapped directly to predictive features, 10 lay within a three-gene window of one, and 13 were functionally linked through the STRING protein-association database. Direct matches were enriched 5.0-fold above random expectation. Interestingly, the model often picked out a single regulatory gene within a larger biosynthetic pathway—porin regulators such as tsx, fadL and nfrA, or individual genes in lipopolysaccharide and capsule clusters—suggesting it captures pathway-level regulation rather than merely listing every component.

Feature interpretation also surfaced both familiar and unexpected biology. Among the top 25 predictive features in the E. coli dataset were elements tied to cell wall biosynthesis, mobile genetic elements, restriction–modification systems and genes of viral origin, alongside eight features with no clear mechanistic link that may represent uncharacterized infection mediators. Network and synteny analyses pointed to candidate roles for putrescine catabolism and regulation of the known phage receptor fepA, hypotheses the authors flag for future experimental work. Defence systems generally carried negative impacts on infection probability, as expected, while anti-defence features trended positive.

The study is candid about its limitations. Cross-genus prediction remains poor, with leave-one-genus-out AUROC values hovering barely above random, indicating that interaction determinants are largely genus-specific or that protein-family features cannot capture conserved mechanisms across deep evolutionary distances. RB-TnSeq validation is restricted to non-essential genes with strong fitness effects, and the authors suggest CRISPR interference screens as a complement for essential genes. Performance also depends on large datasets of confirmed interactions, which are slow and expensive to produce. Nevertheless, the successful merging of Klebsiella datasets from different research groups points toward community-driven, standardized data generation as a viable path forward. As antimicrobial resistance escalates, the authors argue, this framework—released as the open-source GenoPHI package—offers a scalable computational platform for rational phage-therapy design and precision microbiome engineering across clinical, agricultural and industrial settings.

Subject of Research: Machine-learning prediction of strain-level bacteriophage–host interactions from bacterial and phage genome sequences

Article Title: Phylogeny-agnostic strain-level prediction of phage–host interactions from genomes using machine learning

Article References: Noonan, A. J. C., Moriniere, L., Rivera-López, E. O., Patel, K., Pena, M., Svab, M., Kazakov, A., Deutschbauer, A., Dudley, E. G., Mutalik, V. K., & Arkin, A. P. (2026). Phylogeny-agnostic strain-level prediction of phage–host interactions from genomes using machine learning. Nature Microbiology. https://doi.org/10.1038/s41564-026-02482-5

Image Credits: AI Generated

DOI: 10.1038/s41564-026-02482-5

Keywords: bacteriophages, phage therapy, machine learning, phage–host interactions, genomics, antimicrobial resistance, strain-level prediction, RB-TnSeq, phage cocktails, microbiome engineering, CatBoost, protein families

Bacteriophages, or phages, are viruses that specifically infect bacteria. Traditionally, phages studied in detail fall into a few general categories based on their infection mechanisms and host preferences. One such category, flagellotropic phages, is known for the targeted infection of bacteria possessing flagella—long, whip-like appendages used by many bacteria to move through their environments. These phages leverage the mechanical movement and molecular signatures of flagella to locate and anchor themselves to their hosts before initiating infection. The newly described phages PIN1 and PIN2, however, defy this archetypal behavior by infecting bacteria that lack motility altogether.

The research team led by Jati et al. conducted an extensive characterization of PIN1 and PIN2, analyzing their genetic makeup, structural features, and host ranges. Both phages exhibit hallmark molecular and morphological traits aligned with flagellotropic classification. Electron microscopy images reveal tail fibers and attachment structures finely tuned for interacting with bacterial flagella. Yet, intriguingly, their confirmed bacterial hosts are immotile species devoid of functional flagella, suggesting the presence of alternative infection strategies or binding mechanisms that mimic flagella interactions without requiring actual motile appendages.

This paradoxical finding is significant because it challenges the textbook definition of flagellotropic phages and hints at a more nuanced, flexible evolutionary trajectory. Phages and their bacterial hosts are locked in a perpetual arms race, driving the evolution of highly specialized viral recognition systems and bacterial defense mechanisms. PIN1 and PIN2 appear to represent a new evolutionary intermediate or a distinct adaptation pathway, where phages retain structural motifs associated with flagellar targeting but have developed the ability to circumvent the need for bacterial motility.

Moreover, the stability of these phages under a variety of environmental conditions is exceptional. Experiments in the study demonstrated that PIN1 and PIN2 maintain infectivity across a broad range of temperatures, pH levels, and ionic strengths, far surpassing the robustness typically observed in other phages with similar genome sizes. This robust stability is likely a critical factor enabling their persistence in diverse ecological niches where host motility may be restricted or absent.

The implications of PIN1 and PIN2’s stability extend well beyond fundamental science. Phages have been widely explored as alternatives or supplements to antibiotics in combating bacterial infections, particularly in the face of rising antimicrobial resistance. Stability is a prized trait for therapeutic phages, as it enhances shelf life, efficacy, and delivery options. The discovery of highly stable phages capable of infecting immotile bacteria, which often form biofilms or exist in dormant states, suggests novel applications in phage therapy and biotechnology that could improve treatment outcomes.

Detailed genomic analyses uncovered that PIN1 and PIN2 possess gene clusters similar to those found in classical flagellotropic phages. These genes encode for tail fibers, baseplates, and receptor-binding proteins, yet subtle mutations and structural variations imply specialization for alternative host receptors. This suggests a fascinating molecular mimicry or convergent evolution, whereby phage components structurally resemble machinery used for flagellar attachment but engage different, possibly conserved motifs on non-motile bacterial surfaces.

The team also investigated the co-evolutionary dynamics between PIN phages and their hosts through experimental evolution assays. Over successive bacterial generations, no detectable resistance developed against PIN phages, contrasting with many known phage-host pairings where resistance arises rapidly. This may stem from the unique attachment and infection mechanisms employed by these phages, perhaps targeting essential bacterial structures that are less prone to mutational escape due to their critical functional roles, or by engaging multiple receptor sites simultaneously.

In terms of ecology, PIN1 and PIN2’s ability to infect immotile bacteria with typical flagellotropic phage architecture could reflect adaptations to microenvironments where bacterial motility is suppressed or energetically unfavorable. For example, within biofilms, bacteria often downregulate flagellar production, entering sessile modes to optimize resource use and collective resilience. Phages like PIN1 and PIN2 might represent an evolutionary solution to sustain viral propagation under such conditions, adding complexity to microbial community dynamics and virus-mediated horizontal gene transfer.

Structurally, cryo-electron microscopy provided high-resolution visualization of PIN1 and PIN2 virions, illuminating their capsid geometries and tail assembly. Both phages exhibit a contractile tail sheath consistent with the Myoviridae family, known for potent injection mechanisms, which could facilitate penetration of bacterial cell envelopes that differ from motile species’ outer surfaces. The researchers speculate that the structural flexibility inherent in these tails allows targeting of alternative receptors while preserving infection efficiency.

Another notable aspect of the PIN phages is their genomic compactness combined with genetic robustness. The genomes, consisting of linear double-stranded DNA, include a repertoire of genes for DNA replication, structural proteins, and host lysis, but lack accessory genes commonly implicated in host manipulation or motility-specific interactions. This lean genomic design might reflect an optimized infection cycle tailored to stable yet selective host targeting without unnecessary metabolic burden.

Future avenues for research stemming from this study are manifold. Discovering the precise molecular receptors and binding patterns that allow PIN1 and PIN2 to infect immotile bacteria remains a priority. Such information could reveal novel bacterial surface molecules as phage receptors, broadening the spectrum of known host-phage interactions and facilitating the design of phage-based antibacterial agents with finely tuned host specificities.

In addition, exploring the environmental distribution of PIN-like phages could shed light on their ecological roles, prevalence, and influence within natural and clinical settings. Given their high stability, PIN phages may persist in harsh or fluctuating environments, acting as critical agents in bacterial population control and gene exchange, with possible impacts on microbial community structure and function.

The findings also raise fundamental questions about the evolutionary origins of flagellotropic phages. Are PIN1 and PIN2 remnants of ancestral phages that originally co-evolved with motile hosts but subsequently adapted to immotile species? Or do they represent an independent lineage that co-opted flagellotropic features for entirely different infection strategies? Resolving this will require phylogenomic comparisons against a broad database of phage sequences and functional assays to understand adaptation trajectories.

In therapeutic contexts, the unique properties of PIN1 and PIN2 suggest practical benefits. Their exceptional stability could facilitate storage and transportation logistics, overcoming significant hurdles faced by phage therapy products. Moreover, their targeting of immotile bacteria broadens the range of pathogenic species amenable to phage treatment, notably those forming chronic infections where bacteria adopt sessile lifestyles resistant to many antibiotics.

The study by Jati and colleagues thereby not only deepens the mechanistic understanding of phage biology but also provides a blueprint for exploring viral diversity beyond classical paradigms. It highlights the remarkable evolutionary ingenuity of bacteriophages, capable of adapting infection strategies to exploit even seemingly unfavorable host traits, such as the absence of motility structures.

This work resonates with the broader shift in microbiology and virology toward appreciating the vast, largely untapped diversity of viruses in nature. Advances in sequencing, microscopy, and bioinformatics now enable the uncovering of such extraordinary viral phenotypes that redefine established biological concepts and unlock new technological and therapeutic potentials.

As we continue to unravel the complexity of phage-host interactions through studies like this, we edge closer to harnessing these microbial predators effectively for human benefit, ecological management, and biotechnology innovation. PIN1 and PIN2 stand as compelling models for future research aiming to decode the intricate molecular dance between viruses and their bacterial hosts in all their astonishing variety.

Subject of Research: Characterization of bacteriophages PIN1 and PIN2 exhibiting features of flagellotropic phages but infecting immotile bacteria.

Article Title: Highly stable bacteriophages PIN1 and PIN2 have hallmarks of flagellotropic phages but infect immotile bacteria.

Article References:
Jati, A., Li, Y., Mu, A. et al. Highly stable bacteriophages PIN1 and PIN2 have hallmarks of flagellotropic phages but infect immotile bacteria. npj Viruses 3, 56 (2025). https://doi.org/10.1038/s44298-025-00139-4

Image Credits: AI Generated