Secreted mucins are the major component of the mucus layer that protects intestinal epithelial surfaces by blocking excessive interactions with the microbiota. Mucins are complex glycoproteins decorated with over 100 different O-glycans. Some bacteria can utilize mucins and excessive degradation has been associated with disruption of the mucus barrier and inflammation. Despite the importance of mucins, a detailed enzymatic pathway by which gut bacteria degrade colonic mucin O-glycans and the impact of this process on gut colonization are unknown. Here, we identified >100 genes that are expressed by the symbiont Bacteroides thetaiotaomicron during growth on different O-glycan substrates, revealing effects of glycan structure on gene expression. The characterization of 33 glycoside hydrolase enzymes revealed the pathway for colonic O-glycan degradation by this bacterium. In vivo competition experiments show that multiple exo-acting enzymes targeting mucin capping structures are central to gut colonization and may provide targets to inhibit bacterial mucin degradation.

Indole-3-acetic acid (IAA) is a vital plant hormone, yet its natural synthesis is insufficient to meet agricultural demand. The indole-3-acetamide (IAM) pathway offers a promising route for microbial IAA production but suffers from inefficient amidase activity. In this study, we identified and engineered an amidase (RsAD) from Rhodococcus sp. through structural analysis, which revealed a narrow substrate channel limiting IAM access, followed by targeted mutagenesis to generate the optimized mutant RsAD-L447A. This mutant exhibited a 3.1-fold increase in catalytic efficiency (kcat/Km) and enabled complete IAM hydrolysis without the accumulation of intermediates. Coexpression of RsAD-L447A with l-tryptophan monooxygenase established a cascade pathway for IAA synthesis in E. coli. Blocking the competing tnaA-mediated degradation pathway further improved precursor utilization. As a result, IAA production reached 13.3 from 20 g/L l-tryptophan in shake-flask cultures. These findings demonstrate an effective enzyme engineering and metabolic optimization strategy for high-level IAA biosynthesis.

Deciphering gene regulatory networks (GRNs) from single-cell transcriptomics data remains a fundamental challenge in computational biology. It is hindered by data sparsity, high dimensionality, and the lack of scalable, generalizable inference models. To address this, we present GRNFormer, a generalizable graph transformer framework for accurate GRN inference from transcriptomics data across species, cell types, and platforms without requiring cell-type annotations or prior regulatory information.  GRNFormer integrates a transformer-based gene expression encoder (Gene-Transcoder) with a variational graph autoencoder (GraViTAE) employing pairwise attention to jointly learn the representations of genes (nodes) and their co-expression relationships (edges). Leveraging TF-Walker, a transcription factor-anchored subgraph sampling strategy, it effectively captures gene regulatory interactions from either single-cell or bulk RNA-seq datasets. Benchmarking on standard datasets demonstrates that GRNFormer outperforms existing traditional and deep learning state-of-the-art methods in blind evaluations, achieving average Sampled Area Under the Receiver Operating Characteristic Curve (Sampled_AUROC) and Sampled Area Under the Precision-Recall Curve (Sampled_AUPRC) values between 0.90 and 0.98 as well as 0.87–0.98 average Sampled F1 score. The model robustly recovers both known and novel regulatory networks, including pluripotency circuits in human embryonic stem cells (hESCs) and immune cell modules in Peripheral Blood Mononuclear Cells (PBMCs). The architecture enables scalable, biologically interpretable GRN inference across various datasets, cell types, and species, establishing GRNFormer as a robust and transferable tool for network biology.

Transformer-based genomic sequence models represent an emerging frontier in computational biology. Yet, their embeddings have not yet shown the same level of predictive power as natural and protein language models, indicating a gap between current implementations and theoretical promise. Existing benchmarks for DNA language models primarily focus on classifying regulatory elements in eukaryotic genomes, leaving open the fundamental question of whether these models learn sequence-level features across whole genomes. We introduce LAMBDA, a benchmark designed to rigorously evaluate genome language model embeddings through phage-bacteria sequence discrimination across four categories of increasing complexity: probing tasks, fine-tuning assessments, diagnostic tests, and genome-wide prophage detection. Our comprehensive analysis of current genomic language models provides novel insights into the importance of training data quality relative to model size, the need for domain-specific training, and the application of genomic language models for detecting prophage sequences. This benchmark represents a challenging genomic annotation task in the bacterial domain and addresses a key computational problem with direct relevance to microbiology and medicine.

Phage therapy is a re-emerging approach for antimicrobial-resistant bacterial infections. However, the narrow host range of most phages remains a major barrier to the success and wider adoption of phage therapy. Although receptor incompatibility is often assumed to define phage-host specificity, we demonstrate that anti-phage defense systems are major determinants of host range in Staphylococcus aureus. Using a methicillin-resistant S. aureus (MRSA) clinical isolate as a model, we characterized the targeting profiles of its 15 defense systems and, for the first time, generated therapeutic phages that evade the full defense repertoire of a multi-phage-resistant strain. In particular, we show that defense-guided phage recombination is a powerful tool that leverages the modular design of phage genomes to replace targeted with untargeted components. Our holistic approach unveils defense synergies that constrain phage evasion and redundancies that allow the simultaneous evasion of multiple defenses. Finally, we show that an engineered phage cocktail prevents the emergence of phage resistance in the model and a second clinical strain with similar defenses. Our work provides a blueprint for translating our expanding knowledge of defense system identity and mechanism into the rational design of effective, next-generation phage therapeutics.

CRISPR/Cas12a technology, characterized by its distinctive trans-cleavage activity, has evolved beyond its gene-editing function to emerge as a powerful tool for molecular detection. This review systematically delineates its structural foundation and molecular mechanism, with a focus on how the technology converts specific nucleic acid recognition into cascade signal amplification. Its applications span pathogen diagnosis, species identification, food safety, and authentication of traditional Chinese medicines. Through integration with isothermal amplification and multimodal detection platforms, Cas12a has driven molecular diagnostics toward portability, visualization, and quantification. The review further discusses challenges related to sensitivity, quantitative accuracy, crRNA design, and standardization, while outlining future directions through convergence with cutting-edge technologies such as microfluidics and artificial intelligence, offering a forward-looking perspective for the development of next-generation precision biosensing platforms.

Engineering E. coli to coassimilate glucose and CO2 requires rewiring central metabolism so a Calvin-Benson-Bassham (CBB) module can compete with, rather than be overwhelmed by, glycolysis. We implemented a Rubisco-based engineered pathway, i.e., heterologous phosphoribulokinase (PrkA) and ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco), and attenuated glycolysis by (i) CRISPRi repression of gapA (GAPDH) and (ii) deletion of major fermentative redox sinks (ΔldhA Δfrd). Activation of the Rubisco-based engineered pathway not only enabled CO2 fixation but also unexpectedly revived glycolytic throughput, yielding a coordinated “harmony” between the two pathways. The engineered strain (E. coli MZLF/pSLiP, pCCS01) sustained cometabolism, consuming glucose at 170 ± 1 mg L–1 h–1 by 84 h. Flux analysis indicated that 17.0 ± 0.1 mg L–1 h–1, 10% of total glucose uptake, was routed through the Rubisco-based engineered pathway, corresponding to a CO2-fixation rate of 5.0 ± 0.1 mg L–1 h–1. Mechanistically, in the parental E. coli background (ldhA+, frd+) ATP is primarily supplied by glycolysis with redox balance via lactate/succinate formation. With the CBB module active in the engineered context (gapA repressed, ΔldhA Δfrd, prkA+, rbcLS+), pyruvate allocation, ATP from acetate, and NADH reoxidation from ethanol production jointly determine flux partitioning, yielding a distinct fermentation profile. These findings show that successful central-metabolism rewiring must target not only core nodes (e.g., gapA) but also auxiliary redox circuits (ldhA, frd). Equally important is maintaining a moderate activity of the Rubisco-based engineered pathway, which allows restoration of the near-equilibrium state at the GAPDH-repressed G3P/1,3-BPG/3PG node.

DNA methylation (DNAme) is the best studied epigenetic mechanism that plays pivotal role in tissue differentiation and epigenetic disruption has been correlated to diverse disease types (e.g. cancer, metabolic disorders). While various DNAme array platforms have been discovered, data analysis remains a challenging task which often requires in-depth bioinformatic expertise. Here, we developed a user-friendly web-based application for data analysis and visualization that accommodates users ranging from early-career basic/translational researchers to experienced bioinformaticians. CpGene is a web application for analyzing DNA methylation array data. It supports Illumina 450K, EPIC, and EPICv2 methylation array platforms and processes .idat files with integrated preprocessing, normalization, and quality control. Biomarker discovery is available through either classic differential methylation point analysis or machine learning–based feature selection as well as gene enrichment analysis. Results are summarized with clear visualizations, to aid interpretation. By combining these functions in a unified interface, CpGene streamlines methylation analysis and helps identify CpG sites and genes with biological and clinical relevance.

Unicellular microorganisms can make a transition to multicellular states that enhance survival under environmental fluctuations. In bacteria, one of these states is the biofilm, defined by the production of an extracellular matrix. Although biofilm maturation and dispersion have been extensively studied, where and how initial matrix production is induced within a growing population remains largely unknown. Here we show that production of colanic acid, an important matrix component, is initiated around topological defects, where cell orientation mismatches and growth-induced pressure builds up, in bacterial monolayers. Using E. coli reporting mechanically induced production of colanic acid in response to cell contact and deformation, we found matrix production accompanied by out-of-plane growth under agar-pad confinement. Controlling confinement geometry using microfluidic devices dictated the positions of topological defects and thereby localized regions of high matrix production. These findings reveal that the cell orientation patterning spatially organizes mechanical cues to induce matrix production for biofilm initiation of bacteria.