Search bioRxiv⌕ Search

Biology subjects

Wang, B.

Publications and source records attributed to Wang, B..

At least 37 records · Page 2Linked to original sources

Chromosomal mutational signatures of DNA damaging agents at single cell resolution

The chromosomal-scale mutational spectrum of small molecules that interact with DNA has been hard to study at scale, as mutational events are distributed in location and occur in parallel in different cells. Here, we present a framework that pairs phylogenetic ancestry reconstruction with mutational signature decomposition to characterise recent, cell-private copy number alteration (CNA) mutational patterns at single-cell resolution. We used this framework to characterise the cell-wise mutational spectrum of contemporaneous CNAs generated by double-strand-break-inducing chemotherapeutic drugs. We demonstrate that platinum salts, G-quadruplex stabilizers and topoisomerase II inhibitors, although mechanistically distinct, converge on a mutational signature dominated by telomere-bounded copy-number gains and losses. This signature is observed in different genetic backgrounds and in vivo in drug-treated patient-derived xenografts. We also observe a high rate of endogenous telomere-bounded mutational foreground in BRCA1 deficient cells. We show that the single cell genome derived signature exposures are drug dose-dependent, and use this to identify the decay of mutational load after drug withdrawal. We observe that both cisplatin and a G4 binder molecule (CX5461) exhibit foreground mutational signature persistence for at least 3 weeks after drug withdrawal, suggesting that residual effects of exposure may last longer than anticipated. Finally, extending the framework to serially drug-treated patient-derived xenograft (PDX) models, we show that telomere-bounded CNA signature exposure is associated with tumoural response to drug, consistent with loss of mutational activity on the genome after acquired resistance emerges. Together, our results show that our framework applied on scWGS identifies contemporaneous chromosomal mutation patterns induced by small molecules in human tissues.

cancer biology↗

TomatoPGFM: A graph-conditioned foundation model for tomato pangenomes

Most genomic foundation models are pretrained on independent linear assemblies and therefore do not explicitly represent population-level segment sharing or local graph connectivity. We developed TomatoPGFM, a graph-conditioned model pretrained on 54.65 Gb of sequence from 66 tomato (Solanum spp.) accessions. Sequence tokens were conditioned on pangenome node attributes and local adjacency, and the model was optimised using masked language modelling and graph-feature reconstruction. To evaluate model responses to graph-conditioned input, we compared aligned, shuffled and disabled graph inputs in 25,000 windows from the training panel. Sequence-aligned graph input produced lower masked language modelling loss than graph-off at all five curriculum stages in both training-panel strata, while the shuffled perturbation generally yielded intermediate losses. We then assessed sequence-only transfer in Solanum sitiens LA1974 and S. lycopersicum MicroTom, neither of which was used for graph construction or pretraining. Frozen-probe AUROC values for gene-versus-intergenic and coding-sequence-versus-intergenic classification ranged from 0.8489 to 0.9593. TomatoPGFM produced higher AUROC point estimates than DNABERT-2 in all four comparisons. Enabling the zero-feature GraphAdapter pathway with adjacency messaging disabled changed throughput by less than 1% at 512-2,048 positions under the tested configuration. Together, these results show that TomatoPGFM responds consistently to sequence-aligned pangenome context in training-panel sequences and provides informative sequence representations for genic-region classification in accessions excluded from graph construction and pretraining.

bioinformatics↗

Hypothalamic Cell Type Evolution across Vertebrates

The hypothalamus regulates vital functions and is highly conserved among vertebrates occupying diverse niches, but little is known about the evolution of its cell types. Using comparative single-cell transcriptomics across major vertebrate classes, we identified deeply conserved and clade-specific cell types. While homologous cell types express conserved transcription factors and neuropeptides, their regulatory linkages are often rewired across evolution. The paraventricular nucleus largely retains its neuropeptide expression and responds to dehydration with fine cellular resolution from amphibians to mammals, whereas lamina terminalis neurons are specific to terrestrial vertebrates and exhibit gene expression divergence in cell types activated by dehydration and heat. We propose that a conserved set of keystone cell types sustain essential functions across vertebrates, while flexible gene regulatory programs alongside numerous clade-specific cell types offer evolutionary fluidity.

evolutionary biology↗

The cerebellum specializes for language even in the absence of contralateral neocortical inputs

Long considered a structure dedicated primarily to motor control, the cerebellum is now known to contain regions that respond selectively to language. However, how cerebellar language specialization emerges during development remains unknown. The prevailing proposal is that cerebellar functional specialization critically depends on inputs from the contralateral neocortex, received through well-established reciprocal cortico-cerebellar connections. Here, we test this proposal in two individuals with atypical brain development: EG, a right-handed woman who lacks most of her left temporal lobe (presumably from birth) and whose neocortical language network resides in her right hemisphere, and C1, an individual missing the entire left cerebral hemisphere from birth. Using precision functional MRI in EG and a cohort of 74 typically-developing adults, we find that EG's cerebellar language network shows a strong left-hemispheric bias, mirroring the atypical lateralization of language in her cerebral cortex, while preserving canonical topography and response profiles of the language-dominant cerebellar regions. Critically, however, EG's right cerebellar hemisphere also responds to language and even contains a language-selective region despite the absence of language regions in the neocortical left hemisphere. We replicate these cerebellar findings in C1, a more extreme case who is missing the entire left hemisphere. These results show that the cerebellum develops language sensitivity despite the absence of contralateral inputs in the face of early injury. Together, these findings challenge the view that cerebellar specialization critically requires contralateral neocortical inputs, and point instead to some degree of intrinsic neocortex-independent cerebellar organization.

neuroscience↗

Parametric neural control differentiates top neural network models of primate visual cortex

Leading deep neural network encoding models predict visual cortical responses with nearly indistinguishable accuracy, raising the strong inference that these models have converged on the same underlying brain-aligned parameterization of natural image space. Here we demonstrate that this is not the case. We introduce axis-aligned feature accentuation, which converts each model's fitted encoding axis into graded stimulus perturbations that are predicted to parametrically control neural firing within and beyond the natural-image range. We generated over 27,500 controller stimuli from ten leading vision models and presented them to five macaques in closed-loop experiments targeting early, mid-, and high-level visual areas. Despite matched natural image predictivity, models diverged strongly in their ability to control neural firing using accentuated stimuli, revealing that most model encoding axes failed to capture the precise tuning of their corresponding neurons. The two adversarially trained models showed a consistent advantage, though adversarial robustness was only weakly predictive of neural control across other models. Instead, control was better predicted by the spatial frequency structure of the input gradient: the distribution of pixels influencing each encoding axis. Overall, these results establish neural control via axis-aligned feature accentuation as a causal method to assess the alignment between how neurons and models parameterize the visual world.

neuroscience↗

The Maine Coon Cat Harboring the MYBPC3-A31P Mutation: A Genotype-Stratified Phenotypic Characterization of Hypertrophic Cardiomyopathy

Abstract Background: Most animal models of HCM are mouse-based, but the thin interventricular septum in mice makes it difficult to clearly distinguish pathological hypertrophy, which introduces substantial errors and constrains basic HCM research. Cats develop HCM spontaneously, and the common MYBPC3-A31P variant in cats is homologous to human mutations in both genetics and pathology, with a larger body size that makes them suitable as large-animal models. This study examines how heterozygosity or homozygosity for the p.A31P mutation (c.91G>C) in the MYBPC3 gene affects the phenotype and severity of HCM in affected cats, with the aim of establishing an ideal large-animal model for clinical risk stratification and precision diagnosis and treatment of human HCM. Methods: Forty-nine Maine Coon cats were enrolled and stratified into homozygous mutant (HOM, n=8), heterozygous mutant (HET, n=26), and wild-type (WT, n=15) groups. All cats underwent echocardiography, blood pressure measurement, physiological assessment, hematological and biochemical analyses, and cross-species sequence conservation analysis. Results: No significant differences in baseline characteristics including age and body weight were observed among groups (P>0.05). HOM cats exhibited significantly higher left ventricular outflow tract pressure gradients and greater basal septal thickness compared to WT cats (P<0.05), with HET cats showing intermediate values. Analysis of hematological and serum biochemical parameters revealed no evidence of systemic inflammation or hepatic injury. Sequence conservation analysis confirmed that the A31 residue is highly conserved across mammalian species. Conclusions: This study provides a phenotypic characterization of Maine Coon cats carrying the MYBPC3-A31P mutation, revealing marked gene-dose effects on cardiac structure and function, with homozygous individuals exhibiting more severe phenotypic features. This model serves as a large-animal translational platform that not only clarifies genotype-phenotype correlations but also supports risk stratification and precision therapeutic strategies in human HCM. Its spontaneous nature and genetic homology to human disease make it particularly valuable for bridging preclinical findings to clinical application.

genetics↗

SpatialMOC: Accurate reconstruction of spatial multi-omics landscapes through cross-modality prediction

Spatial multi-omics technologies provide unprecedented opportunities to characterize tissue organization by measuring complementary molecular layers within their native spatial context. However, simultaneous profiling of multiple molecular modalities remains technically challenging, limiting the widespread application of spatial multi-omics and leaving most studies reliant on single-modality measurements. Here, we present Spatial Multi-Omics Cross-prediction (Spatial-MOC), a computational framework that reconstructs missing spatial molecular modalities by integrating spatial context with cross-modality representation learning. Across multiple tissues, molecular modalities, developmental stages, sequencing platforms, and degraded datasets, SpatialMOC consistently outper-formed existing computational approaches in bidirectional molecular prediction. Beyond accurate prediction, SpatialMOC faithfully reconstructed tissue architecture, preserved dynamic molecular and regulatory heterogeneity, and recovered biologically meaningful spatial landscapes from technically compromised measurements. Together, these results establish SpatialMOC as a general framework for spatial multi-omics reconstruction, extending the analytical value of existing spatial omics datasets and facilitating comprehensive investigations of tissue organization, development, and disease.

bioinformatics↗

Reversible m6Am methylation of snRNA by FTO controls morphine reward and tolerance without altering analgesia

Mu opioids, such as morphine, are effective analgesics, but their reward and tolerance drive opioid use disorder. A major goal is to achieve analgesia without these harmful effects. Here we show that morphine reward and tolerance require the RNA demethylase FTO. Genetic depletion and pharmacologic inhibition of FTO each reduced morphine reward, measured by conditioned-place preference, and reduced antinociceptive tolerance to morphine and fentanyl, without altering analgesia. Although FTO is known to erase m6A on mRNA, we found no effect of FTO depletion on m6A sites, but markedly increased levels of m6Am on snRNA. The effects of FTO depletion were suppressed in mice that cannot make m6Am, supporting the role of m6Am in morphine reward and tolerance. We show that FTO depletion regulates a gene expression network linked to morphine signaling. FTO inhibitors may therefore provide useful adjuvants to mu opioids in pain management and treatment of opioid use disorder.

neuroscience↗

High prevalence of KPC-3 in carbapenem-resistant Pseudomonas aeruginosa across multiple clonal lineages in China

Background The epidemiology of Klebsiella pneumoniae carbapenemase (KPC)-producing Pseudomonas aeruginosa is rapidly evolving in China. While blaKPC-2 remains the predominant KPC variant in P. aeruginosa, blaKPC-3 has rarely been documented in this pathogen. This study investigated the molecular epidemiology, resistance and virulence characteristics, and plasmid features of blaKPC-3-producing CRPA isolates collected from a tertiary hospital in eastern China. Methods A total of 65 non-duplicate CRPA isolates collected in 2023 were subjected to whole-genome sequencing. Antimicrobial susceptibility testing, phylogenetic analysis, plasmid characterization, conjugation experiments, and virulence assays were performed. Results Among the 65 CRPA isolates, 37 (56.9%) carried blaKPC-3. These blaKPC-3-positive isolates belonged to four sequence types (STs), including ST1076 (62.2%), ST463 (21.6%), ST646 (10.8%), and ST3393 (5.4%). To our knowledge, this is the first report of blaKPC-3 in P. aeruginosa ST463, ST646 and ST3393. All isolates exhibited extensive drug resistance, and 51.8% were resistant to ceftazidime-avibactam. Phylogenetic analysis indicated that blaKPC-3 dissemination was driven by both clonal expansion and horizontal transmission. Comparative genomic analysis identified three kinds of blaKPC-3 -carrying plasmid. A transferable IncP-2 megaplasmid was widely distributed among ST1076, ST646, and ST3393 isolates, whereas non-transferable IncP-10 plasmids were primarily restricted to ST463. The genetic environments and plasmid backbones of blaKPC-3 were highly conserved and closely related to those of blaKPC-2 and its variants, suggesting evolution from pre-existing blaKPC-2-associated plasmids. Virulence analysis demonstrated marked heterogeneity across lineages. ST463 isolates co-harbored exoU and exoS, exhibited enhanced biofilm formation and pyocyanin production, and caused significantly higher mortality in the G. mellonella infection model, indicating a hypervirulent phenotype. Conclusions The blaKPC-3 is becoming an increasingly important determinant of carbapenem resistance in P. aeruginosa in China. The IncP-2 megaplasmid and IncP-10 plasmid derived blaKPC-3 spread across multiple lineages. Continuous genomic surveillance and enhanced infection control measures are urgently needed to prevent its further prevalence in clinical settings.

microbiology↗

A Practice on Antibody Hydrophobic Interaction Chromatography Retention Time Prediction using Pre-Trained Large Language Model Fine-Tuning

Hydrophobicity is a critical property associated with the risk of non-specific binding, and it is commonly assessed using hydrophobic interaction chromatography retention time. Several computational approaches have been developed to predict antibody developability based on pre-trained language models. Such models can be fine-tuned with limited labeled antibody sequences and, in principle, do not require structural information, which is often challenging to obtain. Nevertheless, few studies have achieved strong performance in hydrophobicity prediction without incorporating structural features. Here, we present a case study of fine-tuning the pre-trained model IgBert to predict antibody hydrophobicity. Using Herceptin as a reference, we performed hydrophobic interaction chromatography retention time experiments and generated Herceptin-adjusted datasets. The fine-tuned model achieved a best R2 of 0.916, underscoring the critical role of rigorous data quality control. We also synthesized and validated 20 commercially available antibody sequences, and the results showed that the predicted hydrophobic properties were correctly reflected. Our findings provide practical guidance and highlight considerations for future applications of fine-tuned pre-trained language models in antibody hydrophobicity prediction.

bioinformatics↗

Phagocytic activity of perivascular cell promotes CNS injury pathology

Injury, stroke, and neurological diseases cause persistent accumulation of cellular debris that deteriorates lesion microenvironment and impedes central nervous system (CNS) repair. Debris clearance in the injured CNS has long been attributed primarily to microglia and infiltrating macrophages. Here, we identify perivascular cells as previously unrecognized phagocytes that expand after injury and exhibit robust phagocytic activity. Perivascular cell phagocytosis is conserved across multiple mouse models of CNS injury and human stroke lesions. These cells exhibit key hallmarks of phagocytosis, including LC3-associated phagocytosis for efficient lysosomal degradation. Mechanistically, phosphatidylserine serves as the eat-me signal and Axl mediates myelin debris uptake. Myelin phagocytosis drives perivascular cell proliferation, fibrosis and lesion progression. Genetic deletion of Axl in perivascular cells or pharmacological inhibition with the FDA-approved Axl inhibitor Gilteritinib reduces pathology and improves functional recovery after spinal cord injury. Together, these findings establish Axl-dependent perivascular cell phagocytosis as a therapeutic target for CNS repair.

neuroscience↗

Pharmacological up-regulation of CALM1 restore endothelial cells in intracranial aneurysm pathogenesis

Subarachnoid hemorrhage (SAH) resulted from intracranial aneurysm (IA) rupture is an especially severe form of stroke. Endothelial dysfunction represents the initiating event of IA pathogenesis. Understanding the role of endothelial cells (ECs) underlying formation of IAs is helpful to seek for pharmaceutical treatment strategy. Based on single-cell RNA sequencing, proteomics, and metabolic analysis, we discovered a change in cell population in IA samples, majorly in ECs and macrophages (MPs). Abnormal ECs exhibit senescence and death in IA samples, which is absent in the control arterial samples. Cross-analysis of multi-omics revealed that CALM1, a calcium detector involved in mechanotransduction, is downregulated in the abnormal ECs. CALM1 knockdown leads to senescence and inhibits the proliferation and maturation of ECs under turbulent flow. Through high-throughput virtual screening, this work identified compound ZC04329651 as a potent CALM1 activator in maintaining the stability of endothelial cell junctions and attenuating cellular senescence. Thus, our findings showed compound ZC04329651 up-regulate the expression of CALM1 to restore ECs, which maybe a promising pharmacological treatment strategy for IAs.

cell biology↗

Reduced benzoxazinoid defences favour maize beneficial colonisation by Colletotrichum tofieldiae

O_LISpecialised defence metabolites mediate plant-microbe interactions, but their regulation during beneficial associations in cereals remains poorly understood. Here we investigate how maize (Zea mays) chemical defences are modulated during interaction with the beneficial fungal endophyte Colletotrichum tofieldiae (strain Ct0861) to promote host growth. C_LIO_LITo elucidate the molecular basis of this interaction, we combined transcriptomics, metabolite analyses and functional assays. To assess benzoxazinoids (BXs) roles at the plant-fungus interface, we compared Ct0861 colonisation and growth-promotion in wild-type maize and the BX-deficient bx1::DS mutant, using the pathogen Colletotrichum graminicola (CgM1.001) as reference. C_LIO_LITranscriptomic analyses revealed coordinated regulation of specialised defence pathways during the initial stages of Ct0861 colonisation, featuring systemic induction of terpenoid phytoalexin biosynthesis alongside localised suppression of roots BXs. Metabolomic analyses showed reduced apoplastic MBOA in Ct0861-colonised roots. Toxicity assays evidenced that Ct0861, unlike CgM1.001, is highly sensitive to MBOA. Accordingly, Ct0861 accumulation and growth-promoting effects were significantly enhanced in the bx1::DS mutant without causing disease symptoms. C_LIO_LIOur findings indicate that BXs restrict Ct0861 maize colonisation, where BX deficiency increases fungal biomass and enhances growth promotion. Understanding how crops fine tune specialised metabolism to balance microbial restriction with beneficial accommodation provides valuable insights for sustainable agriculture. C_LI

plant biology↗

Proteomic Profiling Identifies CLDN3 as a Tumor-Selective Therapeutic Target in Small Cell Lung Cancer

Small cell lung cancer (SCLC) remains a highly lethal disease with limited targetable surface antigens beyond delta-like ligand 3 (DLL3). We sought to systematically identify and validate tumor-selective cell-surface targets in relapsed SCLC. To this end, we developed an integrated proteogenomic pipeline combining single-cell RNA sequencing, combinatorial optimization, mass spectrometry-based proteomics, and immunohistochemistry to systematically map the SCLC surfaceome of 49 tumors across 25 patients with relapsed SCLC. Using this approach we found claudin-3 (CLDN3) to be a consistently expressed and tumor-selective antigen, with broader coverage than DLL3, which is the current clinical benchmark. CLDN3 was highly expressed across various treatment states, while maintaining low expression in most nonmalignant tissues. Functional validation using a novel CLDN3-specific monoclonal antibody (ABN501) demonstrated NK cell-mediated cytotoxicity in vitro and tumor regression in vivo, as well as a favorable safety profile. These findings support clinical development of CLDN3-directed therapies and demonstrate the utility of integrated proteogenomic approaches for antigen discovery in solid tumors.

cancer biology↗

NFYA regulates spermatogenesis through multifaceted chromatin interactions

During spermatogenesis, germ cells undergo extensive chromatin remodeling to promote highly diverse and stage-specific transcriptional programs that drive differentiation, meiosis, and spermiogenesis. However, conventional loss-of-function approaches do not allow systematic interrogation of gene function across spermatogenesis in vivo. Here, we introduce a protein-degron strategy coupled with intra-seminiferous tubule injection (dTAG-IS) to enable rapid, simultaneous, and stage-resolved functional investigation of NFYA during spermatogenesis. Acute NFYA-depletion induces apoptosis in differentiating spermatogonia (diff-Spg), pachytene spermatocytes (pSpc), and round spermatids (rStd), revealing an essential requirement across spermatogenesis. Mechanistically, NFYA exhibits stage-specific chromatin-binding patterns and directly regulates distinct transcriptional programs, including cell cycle progression and histone assembly in diff-Spg; chromosome organization and DNA damage response in pSpc; and antioxidative stress programs in rStd. Together, our study underscores dTAG-IS as a versatile strategy for in vivo loss-of-function studies during spermatogenesis and establishes NFYA as a multifaceted regulator that drives stage-specific transcriptional programs to ensure spermatogenesis progression.

developmental biology↗

Seasonal Rebalancing of Assembly Processes and Cross-Domain Network Restructuring in Microbial Communities of a Large Eutrophic Lake

Harmful algal bloom (HAB) succession in eutrophic lakes arises from seasonal rebalancing of selection, dispersal, and drift across interacting microbial domains, yet process attribution remains limited due to the scarcity of cross-domain analyses. Using multi-marker eDNA metabarcoding (16S rRNA, 18S rRNA, 23S rRNA), we show that bacterioplankton, cyanobacteria, phytoplankton, and zooplankton in Lake Taihu undergo coordinated spring-summer transitions, with inter-seasonal turnover far exceeding spatial heterogeneity. Assembly regimes shifted coherently across domains: spring communities were dominated by strong homogeneous selection--especially among cyanobacteria and microeukaryotes--whereas summer communities moved toward drift- or neutrality-like dynamics under bloom-stage mixing that enhanced dispersal. Null- and neutral-model diagnostics supported these trends. Cross-domain co-occurrence networks restructured seasonally, with summer networks becoming more connected and substantially more robust, indicating that bloom conditions foster cohesive, robust interaction structures rather than destabilization. These findings provide evidence that bloom progression restructures aquatic microbiomes through a seasonal rebalance of selection, drift, and dispersal coupled with adaptive strengthening of cross-domain connectivity, providing a process-explicit framework for understanding community stability and ecosystem resilience in eutrophic lakes.

ecology↗

Variants of relative frequency methods for determining transcranial magnetic stimulation motor threshold do not provide accurate estimation of threshold

Relative frequency methods such as five-out-of-ten have been used for a long time to determine motor threshold of transcranial magnetic stimulation (TMS), despite known severe limitations both in accuracy and speed. Variants of relative frequency methods have been developed to improve performance, such as adaptive staircasing and the recent RMT-Finder using binary search. However, they do not solve the fundamental problem underlying the use of relative frequencies to approximate the 50% TMS response rate at threshold. We demonstrate that the probability of false decisions using relative frequency due to at least half the pulses resulting in responses at subthreshold intensities or more than half the pulses resulting in no responses at suprathreshold intensity cannot be sufficiently reduced unless an impractically high pulse count is used, which substantially increases the duration of the thresholding procedure. Simulations in a virtual population of 25,000 subjects reveal that the accuracy of relative frequency methods is not improved by the adaptive staircasing or binary search variants. Although these variants reduce the number of pulses required to reach an estimate compared to the conventional relative frequency method, they are still slower and less accurate than more advanced thresholding methods such as stochastic approximation or maximum likelihood estimation. Therefore, these faster and more accurate methods should become the standard for motor thresholding of TMS, and the use of methods based on relative frequency should be discouraged. HighlightsO_LITMS motor thresholding should not use relative-frequency methods. C_LIO_LIRelative-frequency methods have a high probability of making false decisions. C_LIO_LIVariants of relative-frequency methods do not improve threshold accuracy. C_LIO_LIStochastic approximation or maximum-likelihood estimation are faster and more accurate. C_LI

bioengineering↗

Coevolution-driven reconstruction of multi-taxa siderophore interaction networks reveals topological diversity of microbial exploitation

Microbial communities are shaped by secreted metabolites that mediate ecological interactions, yet predicting these interactions from genomic sequences remains difficult, because the specific recognition between co-functional metabolites (CFMs), such as siderophores, and their receptor proteins (Rec) cannot be inferred from gene annotation alone. This difficulty arises from three factors: the prevalence of Rec-mediated exploitation, the lack of high-accuracy functional annotations, and the absence of genomic co-localization between functionally paired CFM-Rec in Gram-positive bacteria. Here we present the Coevolution-based Interaction Model (CIM), an automated framework that maps specific CFM-Rec pairings directly from uncurated genomic datasets. Using a dynamic joint optimization strategy that accounts for exploitation asymmetry and avoids combinatorial explosion, CIM identifies functional pairings solely through evolutionary covariation. We validated this approach by reconstructing macroscale iron scavenging networks across nine bacterial taxa. Experiments confirmed that CIM can bridge genomic distances exceeding 3 Mb in the Gram-positive genus Rhodococcus to identify unlinked cognate receptors, and can accurately predict cross-utilization by exploiter strains despite substantial receptor sequence heterogeneity in Burkholderiaceae and Rhizobiaceae. Finally, topological analysis of the reconstructed networks shows that siderophore exploitation acts as a universal topological glue, fusing fragmented microbial populations into highly connected communities, and that the exploitability of siderophore production reverses depending on network modularity. CIM thus offers a scalable, sequence-to-ecology approach for predicting interactions mediated by secondary metabolites in microbial communities.

systems biology↗