Colorectal cancer (CRC) is one of the most common cancers and is the second-highest in cancer-related deaths worldwide. The changes in gut homeostasis and microbial dysbiosis lead to the initiation of the tumorigenesis process. Several pathogenic gram-negative bacteria including Fusobacterium nucleatum are the principal contributors to the induction and pathogenesis of CRC. Thus, inhibiting the growth and survival of these pathogens can be a useful intervention strategy. Fibroblast activation protein-2 (Fap2) is an essential membrane protein of F. nucleatum that promotes the adherence of the bacterium to the colon cells, recruitment of immune cells, and induction of tumorigenesis. The present study depicts the design of an in silico vaccine candidate comprising the B-cell and T-cell epitopes of Fap2 for improving cell-mediated and humoral immune responses against CRC. Notably, this vaccine participates in significant protein–protein interactions with human Toll-like receptors, especially with TLR6 reveals, which is most likely to be correlated with its efficacy in eliciting potential immune responses. The immunogenic trait of the designed vaccine was verified by immune simulation approach. The cDNA of the vaccine construct was cloned in silico within the expression vector pET30ax for protein expression. Collectively, the proposed vaccine construct may serve as a promising therapeutic in intervening F. nucleatum-induced human CRC.

The twenty first century has witnessed challenges with antimicrobial resistance (AMR) emerging as a critical global threat. Among its most concerning is antibiotic resistance (ABR), highly linked to beta-lactamases. Among others, Class A beta-lactamases, present significantly with functional diversity, although ESBLs are one of the major concerns. A key defence mechanism in Gram-negative bacteria is the overexpression of ESBLs (Extended spectrum beta-lactamases) which spread across the bacterial population through horizontal gene transfer causes serious nosocomial infections. Since ESBLs have developed to increase their substrate specificity and hydrolyse most cephalosporins, penicillins, and monobactams, research into them is urgently needed.However, despite attempts functional classification, based on sequence identity, fold similarity, the presence or absence of insertions, particularly in loop regions and mode of action, a universally accepted framework remains elusive. Previous studies have broadly categorized Class A beta-lactamases into subclasses A1 and A2, yet the mechanistic intricacies of subclass A2 only as ESBL demand a more nuanced, multilevel analysis, underlying their role in antibiotic resistance. To bridge this knowledge gap, we employed on a comprehensive investigation encompassing sequence, structure, molecular docking, and dynamic analyses to elucidate the mechanistic approach of antibiotic resistance profiles for these two subclasses. Our sequence and structural studies revealed differences, particularly in insertions, structural alignments, and loop regions, including the omega loop and loops near the active site. Molecular docking study demonstrated better binding of the bigger substrate in the active site cavity of A2 subclass representatives. Dynamic analyses further confirmed our findings, employing root mean square deviation (RMSD), root mean square fluctuation (RMSF), flexibility of the extended and omega loops, radius of gyration (Rg), solvent-accessible surface area (SASA), clustering, hydrogen bonding patterns, principal component analysis (PCA), and free energy landscape (FEL). This study provides insights into the molecular distinctions and resistance mechanisms of these subclasses, paving the way for advanced research in antibiotic resistance and strengthening novel therapeutic strategies.

The application of machine learning potentials (MLPs) to accurately simulate enzymatic reactions remains challenging. This is primarily due to the high dimensionality and structural heterogeneity of enzyme systems, as well as the need to incorporate off-equilibrium and transition state conformations into the training data. Integrating MLPs into a quantum mechanical/molecular mechanical (QM/MM) framework through either direct learning or Δ-learning, together with reaction-specific training strategies, can help overcome these limitations. In this work, we present an integrated workflow for MLP/ΔMLP-assisted QM/MM simulations that achieves ab initio (ai) or density functional theory (DFT)-level accuracy in predicting enzyme reaction thermodynamics. Developed within CHARMM and tightly integrated with the mlp_qmmm Python package, the workflow automates training-data generation, data sanitization, MLP/ΔMLP training, and deployment of trained models in QM/MM molecular dynamics (MD) simulations. In addition, a low-overhead interface implemented in CHARMM enables efficient model inference on both CPU and GPU during simulations. The workflow further incorporates an iterative model refinement strategy that systematically improves predictive performance through successive rounds of sampling, high-level labeling, and retraining. The capabilities of the approach are demonstrated using the hydride-transfer reaction catalyzed by four variants of dihydrofolate reductase (DHFR). Compared with conventional ai/DFT-QM/MM simulations, the ΔMLP-assisted approach achieves more than 500-fold acceleration while maintaining subkcal/mol accuracy in predicted reaction free energies and free energy barriers. Iterative refinement further improves the underlying energy and force predictions, leading to more accurate thermodynamic and structural properties obtained from QM/MM simulations with only a modest additional computational cost. Overall, this work establishes a scalable and extensible workflow for the systematic development and iterative refinement of reaction-specific MLP/ΔMLP models, enabling highly accurate and computationally efficient simulations of enzyme-catalyzed reactions.