Accelerated antimicrobial discovery via deep generative models and molecular dynamics simulations

A breakthrough computational strategy for efficient enzymatic digestion of walnut protein to prepare antioxidant peptides

The conventional natural bioactive peptides (NBAPs) enzymatic preparation process is labor-intensive and time-consuming, limiting its application and development. This study proposes an efficient computational strategy (CAE-VD) that integrates a high-accuracy (98.18 %) deep learning model (convolutional auto-encoder, CAE) with virtual digestion (VD) to prepare NBAPs with specified activities. CAE predicts the activity of peptides generated by VD, guiding enzyme selection. CAE-VD identified alkaline protease as the most suitable enzyme for enzymatic preparation of walnut-derived antioxidant peptides compared to pepsin and trypsin, which was confirmed by DPPH and ABTS radical scavenging assays and statistical analyses of peptides. As an emerging computer technology, CAE-VD will apply to other NBAPs. This study demonstrates the efficacy of integrating deep learning with virtual digestion in guiding the enzymatic preparation of NBAPs and highlights the potential of applying advanced computational techniques in the food industry.

Polylysine Derivatives with a Potent Antibacterial Ability for Effectively Treating Methicillin-Resistant Staphylococcus aureus-Induced Endophthalmitis

Bacterial endophthalmitis (BE) is a severe ocular infection that can lead to irreversible blinding ocular disease. When diagnosed with BE, the main treatment approach is empirically administering intravitreal antibiotic injections. However, the excessive use of antibiotics leads to increased drug resistance in pathogens, and the retinal dose-limiting toxicities greatly limit its application in clinic. In this work, we present a series of polylysine derivatives (PLL-n) for the treatment of bacterial endophthalmitis. By precisely adjusting the balance of hydrophilic/hydrophobic, the optimal polymer, PLL-2, demonstrates high efficacy against Staphylococcus aureus (S. aureus), Escherichia coli (E. coli), and various clinically isolated drug-resistant bacteria. The antibacterial mechanism showed that PLL-2 could effectively destroy the bacterial membrane and lead to bacterial death. Due to its unique antibacterial mechanism, PLL-2 exhibits rapid bactericidal kinetics and does not induce bacterial resistance up to 16 generations. More importantly, PLL-2 showed a significant therapeutic effect on a methicillin-resistant S. aureus-induced rat endophthalmitis model, which presents a promising therapeutic approach for managing endophthalmitis.

PepINVENT: generative peptide design beyond natural amino acids

Peptides play a crucial role in drug design and discovery whether as a therapeutic modality or a delivery agent. Non-natural amino acids (NNAAs) have been used to enhance the peptide properties such as binding affinity, plasma stability and permeability. Incorporating novel NNAAs facilitates the design of more effective peptides with improved properties. The generative models used in the field have focused on navigating the peptide sequence space. The sequence space is formed by combinations of a predefined set of amino acids. However, there is still a need for a tool to explore the peptide landscape beyond this enumerated space to unlock and effectively incorporate the de novo design of new amino acids. To thoroughly explore the theoretical chemical space of peptides, we present PepINVENT, a novel generative AI-based tool as an extension to the small molecule molecular design platform, REINVENT. PepINVENT navigates the vast space of natural and non-natural amino acids to propose valid, novel, and diverse peptide designs. The generative model can serve as a central tool for peptide-related tasks, as it was not trained on peptides with specific properties or topologies. The prior was trained to understand the granularity of peptides and to design amino acids for filling the masked positions within a peptide. PepINVENT coupled with reinforcement learning enables the goal-oriented design of peptides using its chemistry-informed generative capabilities. This study demonstrates PepINVENT's ability to explore the peptide space with unique and novel designs and its capacity for property optimization in the context of therapeutically relevant peptides. Our tool can be employed for multi-parameter learning objectives, peptidomimetics, lead optimization, and a variety of other tasks within the peptide domain.

Unveiling the influence of fastest nobel prize winner discovery: alphafold's algorithmic intelligence in medical sciences

CONTEXT

AlphaFold's advanced AI technology has transformed protein structure interpretation. By predicting three-dimensional protein structures from amino acid sequences, AlphaFold has solved the complex protein-folding problem, previously challenging for experimental methods due to numerous possible conformations. Since its inception, AlphaFold has introduced several versions, including AlphaFold2, AlphaFold DB, AlphaFold Multimer, Alpha Missense, and AlphaFold3, each further enhancing protein structure prediction. Remarkably, AlphaFold is recognized as the fastest Nobel Prize winner in science history. This technology has extensive applications, potentially transforming treatment and diagnosis in medical sciences by reducing drug design costs and time, while elucidating structural pathways of human body systems. Numerous studies have demonstrated how AlphaFold aids in understanding health conditions by providing critical information about protein mutations, abnormal protein-protein interactions, and changes in protein dynamics. Researchers have also developed new technologies and pipelines using different versions of AlphaFold to amplify its potential. However, addressing existing limitations is crucial to maximizing AlphaFold's capacity to redefine medical research. This article reviews AlphaFold's impact on five key aspects of medical sciences: protein mutation, protein-protein interaction, molecular dynamics, drug design, and immunotherapy.

METHODS

This review examines the contributions of various AlphaFold versions AlphaFold2, AlphaFold DB, AlphaFold Multimer, Alpha Missense, and AlphaFold3 to protein structure prediction. The methods include an extensive analysis of computational techniques and software used in interpreting and predicting protein structures, emphasizing advances in AI technology and its applications in medical research.

De Novo Design of Highly Stable Binders Targeting Dihydrofolate Reductase in Klebsiella pneumoniae

The study aims to design novel therapeutic inhibitors targeting the DHFR protein of Klebsiella pneumoniae. However, challenges like bacterial resistance to peptides and the limitations of computational models in predicting in vivo behavior must be addressed to refine the design process and improve therapeutic efficacy. This study employed deep learning-based bioinformatics techniques to tackle these issues. The study involved retrieving DHFR protein sequences from Klebsiella strains, aligning them to identify conserved regions, and using deep learning models (OmegaFold, ProteinMPNN) to design de novo inhibitors. Cell-penetrating peptide (CPP) motifs were added to enhance delivery, followed by allergenicity and thermal stability assessments. Molecular docking and dynamics simulations evaluated the binding affinity and stability of the inhibitors with DHFR. A conserved 60-residue region was identified, and 60 de novo binders were generated, resulting in 7200 sequences. After allergenicity prediction and stability testing, 10 sequences with melting points near 70°C were shortlisted. Strong binding affinities were observed, especially for complexes 4OR7-1787 and 4OR7-1811, which remained stable in molecular dynamics simulations, indicating their potential as therapeutic agents. This study designed stable de novo peptides with cell-penetrating properties and strong binding affinity to DHFR. Future steps include in vitro validation to assess their effectiveness in inhibiting DHFR, followed by in vivo studies to evaluate their therapeutic potential and stability. These peptides offer a promising strategy against Klebsiella pneumoniae infections, providing potential alternatives to current antibiotics. Experimental validation will be key to assessing their clinical relevance.

Chemical Innovations of Antimicrobial Polymers for Combating Antimicrobial Resistance

The global rise of antimicrobial resistance (AMR) has rendered many traditional antibiotics ineffective, leading to an urgent need for alternative therapeutic strategies. Antimicrobial polymers, with their ability to rapidly kill bacteria by disrupting or crossing membranes and/or targeting multiple microbial functions without inducing resistance, offer a promising solution. This perspective explores recent innovations in the design and synthesis of antimicrobial polymers, focusing on their chemical motifs, structural derivatives, and their applications in combating systemic and topical infections. We also highlight key challenges in translating these materials from laboratory research to clinical practice, including issues related to the high dose required, bioavailability and stability in systemic infection treatment, and ability to disperse and kill biofilms in localized infection management. By addressing these challenges, antimicrobial polymers could play a crucial role in the development of next-generation therapeutics to combat multidrug-resistant pathogens. This perspective attempts to summarize significant insights for the design and development of advanced antimicrobial polymers to overcome AMR, offering potential pathways to improve clinical outcomes in treating systemic and local infections.

COLOR: A Compositional Linear Operation-Based Representation of Protein Sequences for Identification of Monomer Contributions to Properties

The properties of biological materials like proteins and nucleic acids are largely determined by their primary sequence. Certain segments in the sequence strongly influence specific functions, but identifying these segments, or so-called motifs, is challenging due to the complexity of sequential data. While deep learning (DL) models can accurately capture sequence-property relationships, the degree of nonlinearity in these models limits the assessment of monomer contributions to a property─a critical step in identifying key motifs. Recent advances in explainable AI (XAI) offer attention and gradient-based methods for estimating monomeric contributions. However, these methods are primarily applied to classification tasks, such as binding site identification, where they achieve limited accuracy (40-45%) and rely on qualitative evaluations. To address these limitations, we introduce a DL model with interpretable steps, enabling direct tracing of monomeric contributions. Inspired by the masking technique commonly used in vision and natural language processing domains, we propose a new metric ( I ) for quantitative analysis on datasets mainly containing distinct properties of anticancer peptides (ACP), antimicrobial peptides (AMP), and collagen. Our model exhibits 22% higher explainability than the gradient and attention-based state-of-the-art models, recognizes critical motifs (RRR, RRI, and RSS) that significantly destabilize ACPs, and identifies motifs in AMPs that are 50% more effective in converting non-AMPs to AMPs. These findings highlight the potential of our model in guiding mutation strategies for designing protein-based biomaterials.

Red-Light Triggered CO/ROS Release from Porphyrin-Flavonol Hybrid@PC7A Micelles for Eradicating Escherichia coli

Photo-triggered carbon monoxide (CO) release, mediated by reactive oxygen species (ROS), shows a significant potential in therapeutic applications. However, the existing photosensitizers predominantly function as type II ROS generators. When ROS are present in excess, they are always wasted due to their short half-lives, which limit their ability to travel significant distances and effectively achieve therapeutic outcomes. In this study, the biological function of a single-component molecule, PdHF, is investigated. This molecule is formed by covalently conjugating 3-hydroxyflavone (3-HF) to a palladium(II) tetraphenyltetrabenzoporphyrin (PdTPTBP) photosensitizer. Subsequently, PdHF is loaded into the 2-hexamethyleneimino ethyl methacrylate (C7A)-modified PEG-b-PCL block copolymer (PC7A) to form a PdHF@PC7A micellar system capable of co-releasing CO and ROS under red-light irradiation. CO/ROS co-release can be attributed to the generation of singlet oxygen species that not only oxidize 3-HF to release CO but are concurrently reduced by the tertiary amine, C7A, to form cytotoxic superoxide anions and hydrogen peroxide. In vitro studies on these positively charged micelles validate a high biosafety and excellent antibacterial effects with competent elimination of Gram-negative bacteria, Escherichia coli. Furthermore, evidence of micelle uptake by bacterial cells supports synergistic photodynamic and gas therapy through intracellular CO/ROS co-release.

Screening and preliminary analysis of antimicrobial peptide genes in Octopussinensis

Antimicrobial peptides (AMPs) are small molecular peptides that widely exist in organisms to resist external microbial invasion and play a crucial role in the host's immune defense system. Owing to their functions of efficient broad-spectrum killing of pathogenic microorganisms, immune enhancement, and intestinal health improvement, they have emerged as a focal point in research on the immune defense of aquatic animals in recent years. In this study, a total of 105 putative AMP-derived genes from the genome were screened, and seven candidate AMPs were finally identified by analyzing the differential expression results of the hepatopancreas and the white body transcriptomes combined with machine learning algorithms. Furthermore, the seven synthesized antimicrobial peptides were demonstrated to have good antimicrobial activity. Among them, GAP1 and Big Defensin showed the strongest antibacterial activity. GAP1 and Big Defensin exhibited antibacterial activity against four bacteria (Escherichia coli, Vibro parahaemolyticus, Staphylococcus aureus, and Bacillus subtilis) at low concentrations of 5-10 μM and 3.2-12.9 μM respectively. These data will contribute to the development of AMP-based aquatic drugs.

Isolation, purification and identification of antibacterial peptides from Jinhua ham broth and molecular simulation analyses of their interaction with bacterial porins

The bioactive peptides in Jinhua ham could be released into the broth during cooking. After comparing peptide antibacterial activity from Jinhua ham broth with varying cooking durations, the cooking-2-h broths were selected for further analysis using cation-exchange and reverse-phase-liquid chromatography. The purified peptide sequences were subsequently synthesized and tested for their antibacterial activity. Four peptides (IKKVVKQASEGP, LGRVPRGKKKL, LKGGKKQLQKL, and MDAIKKKMQMLK) were identified with IC50 values for S. typhimurium and S. aureus below 0.4 mg/mL. Molecular docking and dynamics simulations were employed to investigate the interaction between the four antibacterial peptides and the outer membrane protein F (Omp F) of the Salmonella membrane. All four peptides demonstrated binding energies with Omp F lower than -7 kcal/mol. Stability indicators in molecular dynamics showed minimal fluctuations, further confirming the compactness and stability of the peptide-Omp F complexes. Notably, all four peptides altered the conformation of Omp F, thereby enhancing cell membrane permeability.

Cecropin A-Derived Peptide for the Treatment of Osteomyelitis by Inhibiting the Growth of Multidrug-Resistant Bacteria and Eliminating Inflammation

Osteomyelitis poses substantial therapeutic challenges due to the prevalence of multidrug-resistant bacterial infections and associated inflammation. Current treatment regimens often rely on a combination of corticosteroids and antibiotics, which can lead to complications and impede effective bacterial clearance. In this study, we present CADP-10, a Cecropin A-derived peptide, designed to target methicillin-resistant Staphylococcus aureus (MRSA) and multidrug-resistant Escherichia coli (MRE), while simultaneously addressing inflammatory responses. CADP-10 self-assembles into nanobacterial net (NBacN) that selectively identify and bind to bacterial endotoxins (LPS and LTA), disrupting membrane integrity and depolarizing membrane potential, which culminates in bacterial death. Importantly, these NBacN are bound to LPS and LTA from dead bacteria, preventing their engagement with TLR receptors and effectively blocking downstream inflammatory pathways. Our assessments of CADP-10 demonstrate good biosafety in both in vitro and in vivo models. Notably, in a rabbit osteomyelitis model, CADP-10 eliminated MRSA-induced bone infections, mitigated inflammation, and promoted bone tissue regeneration. This research highlights the potential of CADP-10 as a multifunctional antimicrobial agent for the management of infectious inflammatory diseases.

Activity and Safety Optimization of Mesoricin: A Dual-Domain Antifungal Peptide from Mesorhizobium sp

Cryptococcus neoformans infections pose a significant global health threat. This study introduces mesoricin, a novel dual-domain antimicrobial peptide (AMP) scaffold derived from Mesorhizobium sp. identified using an in silico quantitative antifungal activity index (AFI). The peptide structure comprises an α-helix domain, which disrupts microbial membranes but exhibits highly hemolytic activity, and a β-sheet domain, which targets intracellular energy metabolism and resilient pathways. Rational design through α-helix domain removal and AFI-guided mutations yielded a mesoricin variant with enhanced antifungal activity and reduced cytotoxicity. The optimized mesoricin exhibited broad-spectrum antifungal activity against various Cryptococcus and Candida species (MIC 8-16 μg/mL) while maintaining high biosafety (IC50 > 128 μg/mL against human cell lines). Particularly, the variant demonstrated enhanced fungicidal effects at sub-MIC levels and superior biofilm control capabilities compared to the prototype peptide. These findings highlight mesoricins as a promising scaffold for AMP development targeting Cryptococcus infections.

Rapid Response Antimicrobial Peptide Design Strategy Driven by Meta-Learning for Emerging Drug-Resistant Pathogens

Antimicrobial resistance (AMR) presents a critical global health threat requiring urgent intervention. In order to swiftly respond to and control the spread of emerging drug-resistant bacteria at the onset of their proliferation, our aim is to develop a Rapid Response Antimicrobial Peptide (AMP) design strategy (RR-ADS). This framework addresses the challenge of limited pathogen-specific data by achieving robust generalization from minimal samples by meta-learning and reinforcement learning, optimizing both biocompatibility and efficacy against drug-resistant pathogens. Our model has achieved satisfactory results across multiple evaluation metrics, demonstrating the capability to accurately identify and generate AMPs targeted against drug-resistant bacteria with minimal sample sizes. Within 2 weeks, we successfully designed and experimentally verified AMPs against multidrug-resistant Acinetobacter baumannii, achieving a 93.3% positive rate. RR-ADS has effectively demonstrated the potential of meta-learning in tasks involving bioactive peptides and holds promise as an effective alternative measure to address infectious disease public health emergencies.

HMAMP: Designing Highly Potent Antimicrobial Peptides Using a Hypervolume-Driven Multiobjective Deep Generative Model

Antimicrobial peptides (AMPs) have exhibited unprecedented potential as biomaterials in combating multidrug-resistant bacteria, prompting the proposal of many excellent generative models. However, the multiobjective nature of AMP discovery is often overlooked, contributing to the high attrition rate of drug candidates. Here, we propose a novel approach termed hypervolume-driven multiobjective AMP design (HMAMP), which prioritizes the simultaneous optimization of multiattribute AMPs. By synergizing reinforcement learning and a gradient descent algorithm rooted in the hypervolume maximization concept, HMAMP effectively biases generative processes and mitigates the pattern collapse issue. Comparative experiments show that HMAMP significantly outperforms state-of-the-art methods in effectiveness and diversity. A knee-based decision strategy is then employed to fast screen candidates with favorable physicochemical properties, aligning with the enhanced antimicrobial activity and reduced side effects. Molecular visualization further elucidates structural and functional properties of the AMPs. Overall, HMAMP is an effective approach to traverse large and complex exploration spaces to search for idealism-realism trade-off AMPs.

CPPCGM: A Highly Efficient Sequence-Based Tool for Simultaneously Identifying and Generating Cell-Penetrating Peptides

Cell-penetrating peptides (CPPs) are usually short oligopeptides with 5-30 amino acid residues. CPPs have been proven as important drug delivery vehicles into cells through different mechanisms, demonstrating their potential as therapeutic candidates. However, experimental screening and synthesis of CPPs could be time-consuming and expensive. Recently, numerous attempts have been made to develop computational methods as a cost-effective way for screening a number of potential CPP candidates. Despite significant advancements, current methods exhibit limited feature representation capabilities, thereby constraining the potential for further performance enhancements. In this study, we developed a deep learning framework called CPPCGM, which uses protein language models (PLMs) to identify and generate novel CPPs. There are two separate blocks in this framework: CPPClassifier and CPPGenerator. The former utilizes three pretrained models for simple voting, thereby accurately categorizing CPPs and non-CPPs. The latter, similar to a generative adversarial network, including a discriminator and a generator, generates peptides that are not present in the training data set. Our proposed CPPCGM has achieved remarkably high Matthews correlation coefficient scores of 0.876, 0.923, and 0.664 on three data sets based on the classification results. Compared with the state-of-the-art methods, the performance of our method is significantly improved. The results also demonstrated the generating potential of CPPCGM through qualitative and quantitative evaluation of the generated samples. Significantly, using PLM-based methods can optimize peptides for biochemical functions, benefiting drug delivery and biomedical applications. Materials related are publicly available at https://github.com/QiufenChen/CPPCGM.

A beginner's approach to deep learning applied to VS and MD techniques

It has become impossible to imagine the fields of biochemistry and medicinal chemistry without computational chemistry and molecular modelling techniques. In many steps of the drug development process in silico methods have become indispensable. Virtual screening (VS) can tremendously expedite the early discovery phase, whilst the use of molecular dynamics (MD) simulations forms a powerful additional tool to in vitro methods throughout the entire drug discovery process. In the field of biochemistry, MD has also become a compelling method for studying biophysical systems (e.g., protein folding) complementary to experimental techniques. However, both VS and MD come with their own limitations and methodological difficulties, from hardware limitations to restrictions in algorithmic capabilities. One solution to overcoming these difficulties lies in the field of machine learning (ML), and more specifically deep learning (DL). There are many ways in which DL can be applied to these molecular modelling techniques to achieve more accurate results in a more efficient manner or expedite the data analysis of the acquired results. Despite steadily increasing interest in DL amidst computational chemists, knowledge is still limited and scattered over different resources. This review is aimed at computational chemists with knowledge of molecular modelling, who wish to possibly integrate DL approaches in their research and already have a basic understanding of the fundamentals of DL. This review focusses on a survey of recent applications of DL in molecular modelling techniques. The different sections are logically subdivided, based on where DL is integrated in the research: (1) for the improvement of VS workflows, (2) for the improvement of certain workflows in MD simulations, (3) for aiding in the calculations of interatomic forces, or (4) for data analysis of MD trajectories. It will become clear that DL has the capacity to completely transform the way molecular modelling is carried out.

Antimicrobial Peptides: Mechanism, Expressions, and Optimization Strategies

Antimicrobial peptides (AMPs) are favoured because of their broad-spectrum antimicrobial properties and because they do not easily develop microbial resistance. However, the low yield and difficult extraction processes of AMPs have become bottlenecks in large-scale industrial applications and scientific research. Microbial recombinant production may be the most economical and effective method of obtaining AMPs in large quantities. In this paper, we review the mechanism, summarize the current status of microbial recombinant production, and focus on strategies to improve the yield and activity of AMPs, in order to provide a reference for their large-scale production.

A data-efficient strategy for building high-performing medical foundation models

Foundation models are pretrained on massive datasets. However, collecting medical datasets is expensive and time-consuming, and raises privacy concerns. Here we show that synthetic data generated via conditioning with disease labels can be leveraged for building high-performing medical foundation models. We pretrained a retinal foundation model, first with approximately one million synthetic retinal images with physiological structures and feature distribution consistent with real counterparts, and then with only 16.7% of the 904,170 real-world colour fundus photography images required in a recently reported retinal foundation model (RETFound). The data-efficient model performed as well or better than RETFound across nine public datasets and four diagnostic tasks; and for diabetic-retinopathy grading, it used only 40% of the expert-annotated training data used by RETFound. We also support the generalizability of the data-efficient strategy by building a classifier for the detection of tuberculosis on chest X-ray images. The text-conditioned generation of synthetic data may enhance the performance and generalization of medical foundation models.

Deep learning-based design and screening of benzimidazole-pyrazine derivatives as adenosine A2B receptor antagonists

The Adenosine A2B receptor (A2BAR) is considered a novel potential target for the immunotherapy of cancer, and A2BAR antagonists have an inhibitory effect on tumor growth, proliferation, and metastasis. In our previous studies, we identified a class of benzimidazole-pyrazine scaffolds whose derivatives exhibited the antagonistic effect but lacked subtype selectivity towards A2BAR. In this work, we developed a scaffold-based protocol that incorporates a deep generative model and multilayer virtual screening to design benzimidazole-pyrazine derivatives as potential selective A2BAR antagonists. By utilizing a generative model with reported A2BAR antagonists as the training set, we built up a scaffold-focused library of benzimidazole-pyrazine derivatives and processed a virtual screening protocol to discover potential A2BAR antagonists. Finally, five molecules with different Bemis-Murcko scaffolds were identified and exhibited higher binding free energies than the reference molecule 12o. Further computational analysis revealed that the 3-benzyl derivative ABA-1266 presented high selectivity toward A2BAR and showed preferred draggability, providing future potent development of selective A2BAR antagonists.

Antimicrobial Peptides and Their Mimetics: Promising Candidates of Next-Generation Therapeutic Agents Combating Multidrug-Resistant Bacteria

The increasing morbidity and mortality caused by multidrug-resistant bacteria alerts human beings to the fact that conventional antibiotics are no longer reliable and effective alternatives are imperatively needed. Owing to wide range of sources, diverse structures, and unique mode of action, antimicrobial peptides have been highly anticipated and extensively studied in recent years. Besides, the integration of artificial intelligence helps researchers gain access to the vast unexplored chemical space, which opens more opportunities for the optimization and design of novel structures. Moreover, Due to advances in chemistry and synthetic biology, researchers have also begun to focus on the potential of chemical mimetics of antimicrobial peptides. In this review, a comprehensive discussion about natural and synthesized antimicrobial peptides as well as their chemical mimetics is made, so as to provide a comprehensive summary of this field and inspire follow-up research.

Deep Learning Combined with Quantitative Structure‒Activity Relationship Accelerates De Novo Design of Antifungal Peptides

Novel antifungal drugs that evade resistance are urgently needed for Candida infections. Antifungal peptides (AFPs) are potential candidates due to their specific mechanism of action, which makes them less prone to developing drug resistance. An AFP de novo design method, Deep Learning-Quantitative Structure‒Activity Relationship Empirical Screening (DL-QSARES), is developed by integrating deep learning and quantitative structure‒activity relationship  empirical screening. After generating candidate AFPs (c_AFPs) through the recombination of dominant amino acids and dipeptide compositions, natural language processing models are utilized and quantitative structure‒activity relationship (QSAR) approaches based on physicochemical properties to screen for promising c_AFPs. Forty-nine promising c_AFPs are screened, and their minimum inhibitory concentrations (MICs) against C. albicans are determined to be 3.9-125 µg mL-1, of which four leading c_AFPs (AFP-8, -10, -11, and -13) has MICs of <10 µg mL-1 against the four tested pathogenic fungi, and AFP-13 has excellent therapeutic efficacy in the animal model.

Rationally designed highly amphipathic antimicrobial peptides demonstrating superior bacterial selectivity relative to the corresponding α-helix peptide

De novo design of antimicrobial peptides is a pivotal strategy for developing new antibacterial agents, leveraging its rapid and efficient nature. (XXYY)n, where X represents cationic residues, Y denotes hydrophobic residues, and n varies from 2 to 4, is a classical α-helix template. Based on which, numerous antimicrobial peptides have been synthesized. Herein, we hypothesize that the amphipathy of this type of α-helix template can be further enhanced based on the principles of α-helical protein folding, characterized by a rotation occurring every 3.6 amino acid residues, and propose the highly amphipathic template XXYYXXYXXYYX (where X represents cationic residues and Y denotes hydrophobic residues). Accordingly, the amino acid composition and arrangement of the α-helix peptide (RRWF)3 are adjusted, yielding the highly amphipathic counterpart H-R (RRWFRRWRRWFR). The structure-activity relationship of which is further explored through the substitution of residues at positions 8 and 12. Notably, the highly amphipathic peptides exhibit enhanced antimicrobial activity and reduced hemolytic toxicity compared to (RRWF)3, resulting in superior bacterial selectivity. The most highly amphipathic peptide, H-R, demonstrates potent activity against biofilms and multidrug-resistant bacteria, low propensity for resistance, and high safety and effectiveness in vivo. The antibacterial mechanisms of H-R are also preliminarily investigated in this study. As noted, H-R represents a promising antimicrobial candidate for addressing infections associated with drug-resistant bacteria.

Exploring the repository of de novo-designed bifunctional antimicrobial peptides through deep learning

Antimicrobial peptides (AMPs) are attractive candidates to combat antibiotic resistance for their capability to target biomembranes and restrict a wide range of pathogens. It is a daunting challenge to discover novel AMPs due to their sparse distributions in a vast peptide universe, especially for peptides that demonstrate potencies for both bacterial membranes and viral envelopes. Here, we establish a de novo AMP design framework by bridging a deep generative module and a graph-encoding activity regressor. The generative module learns hidden 'grammars' of AMP features and produces candidates sequentially pass antimicrobial predictor and antiviral classifiers. We discovered 16 bifunctional AMPs and experimentally validated their abilities to inhibit a spectrum of pathogens in vitro and in animal models. Notably, P076 is a highly potent bactericide with the minimal inhibitory concentration of 0.21 μM against multidrug-resistant Acinetobacter baumannii, while P002 broadly inhibits five enveloped viruses. Our study provides feasible means to uncover the sequences that simultaneously encode antimicrobial and antiviral activities, thus bolstering the function spectra of AMPs to combat a wide range of drug-resistant infections.

Discovery of antimicrobial peptides with notable antibacterial potency by an LLM-based foundation model

Large language models (LLMs) have shown remarkable advancements in chemistry and biomedical research, acting as versatile foundation models for various tasks. We introduce AMP-Designer, an LLM-based approach, for swiftly designing antimicrobial peptides (AMPs) with desired properties. Within 11 days, AMP-Designer achieved the de novo design of 18 AMPs with broad-spectrum activity against Gram-negative bacteria. In vitro validation revealed a 94.4% success rate, with two candidates demonstrating exceptional antibacterial efficacy, minimal hemotoxicity, stability in human plasma, and low potential to induce resistance, as evidenced by significant bacterial load reduction in murine lung infection experiments. The entire process, from design to validation, concluded in 48 days. AMP-Designer excels in creating AMPs targeting specific strains despite limited data availability, with a top candidate displaying a minimum inhibitory concentration of 2.0 micrograms per milliliter against Propionibacterium acnes. Integrating advanced machine learning techniques, AMP-Designer demonstrates remarkable efficiency, paving the way for innovative solutions to antibiotic resistance.

Deep Learning for Antimicrobial Peptides: Computational Models and Databases

Antimicrobial peptides are a promising strategy to combat antimicrobial resistance. However, the experimental discovery of antimicrobial peptides is both time-consuming and laborious. In recent years, the development of computational technologies (especially deep learning) has provided new opportunities for antimicrobial peptide prediction. Various computational models have been proposed to predict antimicrobial peptide. In this review, we focus on deep learning models for antimicrobial peptide prediction. We first collected and summarized available data resources for antimicrobial peptides. Subsequently, we summarized existing deep learning models for antimicrobial peptides and discussed their limitations and challenges. This study aims to help computational biologists design better deep learning models for antimicrobial peptide prediction.

Conotoxins: Classification, Prediction, and Future Directions in Bioinformatics

Conotoxins, a diverse family of disulfide-rich peptides derived from the venom of Conus species, have gained prominence in biomedical research due to their highly specific interactions with ion channels, receptors, and neurotransmitter systems. Their pharmacological properties make them valuable molecular tools and promising candidates for therapeutic development. However, traditional conotoxin classification and functional characterization remain labor-intensive, necessitating the increasing adoption of computational approaches. In particular, machine learning (ML) techniques have facilitated advancements in sequence-based classification, functional prediction, and de novo peptide design. This review explores recent progress in applying ML and deep learning (DL) to conotoxin research, comparing key databases, feature extraction techniques, and classification models. Additionally, we discuss future research directions, emphasizing the integration of multimodal data and the refinement of predictive frameworks to enhance therapeutic discovery.

Artificial intelligence using a latent diffusion model enables the generation of diverse and potent antimicrobial peptides

Artificial intelligence holds great promise for the design of antimicrobial peptides (AMPs); however, current models face limitations in generating AMPs with sufficient novelty and diversity, and they are rarely applied to the generation of antifungal peptides. Here, we develop an alternative pipeline grounded in a diffusion model and molecular dynamics for the de novo design of AMPs. The peptides generated by our pipeline have lower similarity and identity than those of other reported methodologies. Among the 40 peptides synthesized for an experimental validation, 25 exhibit either antibacterial or antifungal activity. AMP-29 shows selective antifungal activity against Candida glabrata and in vivo antifungal efficacy in a murine skin infection model. AMP-24 exhibits potent in vitro activity against Gram-negative bacteria and in vivo efficacy against both skin and lung Acinetobacter baumannii infection models. The proposed approach offers a pipeline for designing diverse AMPs to counteract the threat of antibiotic resistance.

Data-driven parametrization of molecular mechanics force fields for expansive chemical space coverage

A force field is a critical component in molecular dynamics simulations for computational drug discovery. It must achieve high accuracy within the constraints of molecular mechanics' (MM) limited functional forms, which offers high computational efficiency. With the rapid expansion of synthetically accessible chemical space, traditional look-up table approaches face significant challenges. In this study, we address this issue using a modern data-driven approach, developing ByteFF, an Amber-compatible force field for drug-like molecules. To create ByteFF, we generated an expansive and highly diverse molecular dataset at the B3LYP-D3(BJ)/DZVP level of theory. This dataset includes 2.4 million optimized molecular fragment geometries with analytical Hessian matrices, along with 3.2 million torsion profiles. We then trained an edge-augmented, symmetry-preserving molecular graph neural network (GNN) on this dataset, employing a carefully optimized training strategy. Our model predicts all bonded and non-bonded MM force field parameters for drug-like molecules simultaneously across a broad chemical space. ByteFF demonstrates state-of-the-art performance on various benchmark datasets, excelling in predicting relaxed geometries, torsional energy profiles, and conformational energies and forces. Its exceptional accuracy and expansive chemical space coverage make ByteFF a valuable tool for multiple stages of computational drug discovery.

Explainable deep learning and virtual evolution identifies antimicrobial peptides with activity against multidrug-resistant human pathogens

Artificial intelligence (AI) is a promising approach to identify new antimicrobial compounds in diverse microbial species. Here we developed an AI-based, explainable deep learning model, EvoGradient, that predicts the potency of antimicrobial peptides (AMPs) and virtually modifies peptide sequences to produce more potent AMPs, akin to in silico directed evolution. We applied this model to peptides encoded in low-abundance human oral bacteria, resulting in the virtual evolution of 32 peptides into potent AMPs. Of these, the 6 most effective were synthesized and tested against multidrug-resistant pathogens and demonstrated activity against carbapenem-resistant species Escherichia coli, Klebsiella pneumoniae and Acinetobacter baumannii, and vancomycin-resistant Enterococcus faecium. The most potent AMP, pep-19-mod, was validated in vivo, achieving over 95% reduction in bacterial loads in mouse models of thigh infection through both systemic and local administration. Our approach advances the automatic identification and optimization of AMPs.

Advances of deep Neural Networks (DNNs) in the development of peptide drugs

Peptides are able to bind to difficult disease targets with high potency and specificity, providing great opportunities to meet unmet medical requirements. Nevertheless, the unique features of peptides, such as their small size, high structural flexibility, and scarce data availability, bring extra challenges to the design process. Firstly, this review sums up the application of peptide drugs in treating diseases. Then, the review probes into the advantages of Deep Neural Networks (DNNs) in predicting and designing peptide structures. DNNs have demonstrated remarkable capabilities in structural prediction, enabling accurate three-dimensional modeling of peptide drugs through models like AlphaFold and its successors. Finally, the review deliberates on the challenges and coping strategies of DNNs in the development of peptide drugs, along with future research directions. Future research directions focus on further improving the accuracy and efficiency of DNN-based peptide drug design, exploring novel applications of peptide drugs, and accelerating their clinical translation. With continuous advancements in technology and data accumulation, DNNs are poised to play an increasingly crucial role in the field of peptide drug development.

High-Accuracy Identification and Structure-Activity Analysis of Antioxidant Peptides via Deep Learning and Quantum Chemistry

Antioxidant peptides (AOPs) hold great promise for mitigating oxidative-stress-related diseases, but their discovery is hindered by inefficient and time-consuming traditional methods. To address this, we developed an innovative framework combining machine learning and quantum chemistry to accelerate AOP identification and analyze structure-activity relationships. A Bi-LSTM-based model, AOPP, achieved superior performance with accuracies of 0.9043 and 0.9267, precisions of 0.9767 and 0.9848, and Matthews correlation coefficients (MCCs) of 0.818 and 0.859 on two data sets, outperforming existing methods. Compared with XGBoost and LightGBM, AOPP demonstrated a 4.67% improvement in accuracy. Feature fusion significantly enhanced classification, as validated by UMAP visualization. Experimental validation of ten peptides confirmed the antioxidant activity, with LLA exhibiting the highest DPPH and ABTS scavenging rates (0.108 and 0.437 mmol/g, respectively). Quantum chemical calculations identified LLA's lowest HOMO-LUMO gap (ΔE = 0.26 eV) and C3-H26 as the key active site contributing to its superior antioxidant potential. This study highlights the synergy of machine learning and quantum chemistry, offering an efficient framework for AOP discovery with broad applications in therapeutics and functional foods.

Leveraging large language models for peptide antibiotic design

Large language models (LLMs) have significantly impacted various domains of our society, including recent applications in complex fields such as biology and chemistry. These models, built on sophisticated neural network architectures and trained on extensive datasets, are powerful tools for designing, optimizing, and generating molecules. This review explores the role of LLMs in discovering and designing antibiotics, focusing on peptide molecules. We highlight advancements in drug design and outline the challenges of applying LLMs in these areas.

Molecular Dynamics Simulations of Structurally Nanoengineered Antimicrobial Peptide Polymers Interacting with Bacterial Cell Membranes

Multidrug resistance (MDR) to conventional antibiotics is one of the most urgent global health threats, necessitating the development of effective and biocompatible antimicrobial agents that are less inclined to provoke resistance. Structurally nanoengineered antimicrobial peptide polymers (SNAPPs) are a novel and promising class of such alternatives. These star-shaped polymers are made of a dendritic core with multiple arms made of copeptides with varying amino acid sequences. Through a comprehensive set of in vivo experiments, we previously showed that SNAPPs with arms made of random blocks of lysine (K) and valine (V) residues exhibit sub-μM efficacy against Gram-negative and Gram-positive bacteria tested. Cryo-TEM images suggested pore formation by a SNAPP with random block copeptide arms as one of their modes of actions. However, the molecular mechanisms responsible for this mode of action of SNAPPs are not fully understood. To address this gap, we employed an atomistic molecular dynamics simulation technique to investigate the influence of three different sequences of amino acids, namely (1) alt-block KKV, (2) ran-block, and (3) diblock motifs on the secondary structure of their arms and SNAPP's overall configuration as well as their interactions with lipid bilayer. We, for the first time, identified a step-by-step mechanism through which alt-block and random SNAPPs interact with lipid bilayer and lead to "pore formation", hence, cell death. These insights provide a strong foundation for further optimization of the chemical structure of SNAPPs for maximum performance against MDR bacteria, therefore offering a promising avenue for addressing antibiotic resistance and the development of effective antibacterial agents.

Challenges and applications of artificial intelligence in infectious diseases and antimicrobial resistance

Artificial intelligence (AI) has transformed infectious disease control, enhancing rapid diagnosis and antibiotic discovery. While conventional tests delay diagnosis, AI-driven methods like machine learning and deep learning assist in pathogen detection, resistance prediction, and drug discovery. These tools improve antibiotic stewardship and identify effective compounds such as antimicrobial peptides and small molecules. This review explores AI applications in diagnostics, therapy, and drug discovery, emphasizing both strengths and areas needing improvement.

DRAMP 4.0: an open-access data repository dedicated to the clinical translation of antimicrobial peptides

Antimicrobial peptides (AMPs) are potential candidates for treating multidrug-resistant bacterial infections, yet only a small number of them have progressed into clinical trials. The main challenges include the poor stability and hemolytic/cytotoxic properties of AMPs. Considering this, in the update of the Data Repository of Antimicrobial Peptides (DRAMP), a new annotation on serum and protease stability is added, and special efforts were made to update the hemolytic/cytotoxic information of AMPs. The DRAMP 4.0 currently holds 30 260 entries (8 001 newly added), consisting of 11 612 general entries, 17 886 patent entries, 96 clinical entries, 377 specific entries, 110 entries with stability data, and 179 expanded entries. A total of 2891 entries possess experimentally determined hemolytic activity information, while 2674 entries contain cytotoxicity data by experimental validation. The update also covers new annotations, statistics, categories, functions, and download links. DRAMP is available online at http://dramp.cpu-bioinfor.org/.

Functional antimicrobial peptide-loaded 3D scaffolds for infected bone defect treatment with AI and multidimensional printing

Infection is the most prevalent complication of fractures, particularly in open fractures, and often leads to severe consequences. The emergence of bacterial resistance has significantly exacerbated the burden of infection in clinical practice, making infection control a significant treatment challenge for infectious bone defects. The implantation of a structural stent is necessary to treat large bone defects despite the increased risk of infection. Therefore, there is a need for the development of novel antibacterial therapies. The advancement in antibacterial biomaterials and new antimicrobial drugs offers fresh perspectives on antibacterial treatment. Although antimicrobial 3D scaffolds are currently under intense research focus, relying solely on material properties or antibiotic action remains insufficient. Antimicrobial peptides (AMPs) are one of the most promising new antibacterial therapy approaches. This review discusses the underlying mechanisms behind infectious bone defects and presents research findings on antimicrobial peptides, specifically emphasizing their mechanisms and optimization strategies. We also explore the potential prospects of utilizing antimicrobial peptides in treating infectious bone defects. Furthermore, we propose that artificial intelligence (AI) algorithms can be utilized for predicting the pharmacokinetic properties of AMPs, including absorption, distribution, metabolism, and excretion, and by combining information from genomics, proteomics, metabolomics, and clinical studies with computational models driven by machine learning algorithms, scientists can gain a comprehensive understanding of AMPs' mechanisms of action, therapeutic potential, and optimizing treatment strategies tailored to individual patients, and through interdisciplinary collaborations between computer scientists, biologists, and clinicians, the full potential of AI in accelerating the discovery and development of novel AMPs will be realized. Besides, with the continuous advancements in 3D/4D/5D/6D technology and its integration into bone scaffold materials, we anticipate remarkable progress in the field of regenerative medicine. This review summarizes relevant research on the optimal future for the treatment of infectious bone defects, provides guidance for future novel treatment strategies combining multi-dimensional printing with new antimicrobial agents, and provides a novel and effective solution to the current challenges in the field of bone regeneration.

Superficial Fungal Infections and Artificial Intelligence: A Review on Current Advances and Opportunities: REVISION

BACKGROUND

Superficial fungal infections are among the most common infections in world, they mainly affect skin, nails and scalp without further invasion. Superficial fungal diseases are conventionally diagnosed with direct microscopy, fungal culture or histopathology, treated with topical or systemic antifungal agents and prevented in immunocompetent patients by improving personal hygiene. However, conventional diagnostic tests can be time-consuming, also treatment can be insufficient or ineffective and prevention can prove to be demanding. Artificial Intelligence (AI) refers to a digital system having an intelligence akin to a human being. The concept of AI has existed since 1956, but hasn't been practicalised until recently. AI has revolutionised medical research in the recent years, promising to influence almost all specialties of medicine.

OBJECTIVE

An increasing number of articles have been published about the usage of AI in cutaneous mycoses.

METHODS

In this review, the key findings of articles about utilisation of AI in diagnosis, treatment and prevention of superficial fungal infections are summarised. Moreover, the need for more research and development is highlighted.

RESULTS

Fifty-four studies were reviewed. Onychomycosis was the most researched superficial fungal infection. AI can be used diagnosing fungi in macroscopic and microscopic images and classify them to some extent. AI can be a tool and be used as a part of something bigger to diagnose superficial mycoses.

CONCLUSION

AI can be used in all three steps of diagnosing, treating and preventing. AI can be a tool complementary to the clinician's skills and laboratory results.

A review on the screening methods for the discovery of natural antimicrobial peptides

Natural antimicrobial peptides (AMPs) are promising candidates for the development of a new generation of antimicrobials to combat antibiotic-resistant pathogens. They have found extensive applications in the fields of medicine, food, and agriculture. However, efficiently screening AMPs from natural sources poses several challenges, including low efficiency and high antibiotic resistance. This review focuses on the action mechanisms of AMPs, both through membrane and non-membrane routes. We thoroughly examine various highly efficient AMP screening methods, including whole-bacterial adsorption binding, cell membrane chromatography (CMC), phospholipid membrane chromatography binding, membrane-mediated capillary electrophoresis (CE), colorimetric assays, thin layer chromatography (TLC), fluorescence-based screening, genetic sequencing-based analysis, computational mining of AMP databases, and virtual screening methods. Additionally, we discuss potential developmental applications for enhancing the efficiency of AMP discovery. This review provides a comprehensive framework for identifying AMPs within complex natural product systems.

Specifically targeted antimicrobial peptides synergize with bacterial-entrapping peptide against systemic MRSA infections

INTRODUCTION

The design of precision antimicrobials aims to personalize the treatment of drug-resistant bacterial infections and avoid host microbiota dysbiosis.

OBJECTIVES

This study aimed to propose an efficient de novo design strategy to obtain specifically targeted antimicrobial peptides (STAMPs) against methicillin-resistant Staphylococcus aureus (MRSA).

METHODS

We evaluated three strategies designed to increase the selectivity of antimicrobial peptides (AMPs) for MRSA and mainly adopted an advanced hybrid peptide strategy. First, we proposed a traversal design to generate sequences, and constructed machine learning models to predict the anti-S. aureus activity levels of unknown peptides. Subsequently, six peptides were predicted to have high activity, among which, a broad-spectrum AMP (P18) was selected. Finally, the two targeting peptides from phage display libraries or lysostaphin were used to confer specific anti-S. aureus activity to P18. STAMPs were further screened out from hybrid peptides based on their in vitro and in vivo activities.

RESULTS

An advanced hybrid peptide strategy can enhance the specific and targeted properties of broad-spectrum AMPs. Among 25 assessed peptides, 10 passed in vitro tests, and two peptides containing one bacterial-entrapping peptide (BEP) and one STAMP passed an in vivo test. The lead STAMP (P18E6) disrupted MRSA cell walls and membranes, eliminated established biofilms, and exhibited desirable biocompatibility, systemic distribution and efficacy, and immunomodulatory activity in vivo. Furthermore, a bacterial-entrapping peptide (BEP, SP5) modified from P18, self-assembled into nanonetworks and rapidly entrapped MRSA. SP5 synergized with P18E6 to enhance antibacterial activity in vitro and reduced systemic MRSA infections.

CONCLUSIONS

This strategy may aid in the design of STAMPs against drug-resistant strains, and BEPs can serve as powerful STAMP adjuvants.

Integrating biophysical modeling, quantum computing, and AI to discover plastic-binding peptides that combat microplastic pollution

Methods are needed to mitigate microplastic (MP) pollution to minimize their harm to the environment and human health. Given the ability of polypeptides to adsorb strongly to materials of micro- or nanometer size, plastic-binding peptides (PBPs) could help create bio-based tools for detecting, filtering, or degrading MNP pollution. However, the development of such tools is prevented by the lack of PBPs. In this work, we discover and evaluate PBPs for several common plastics by combining biophysical modeling, molecular dynamics (MD), quantum computing, and reinforcement learning. We frame peptide affinity for a given plastic through a Potts model that is a function of the amino acid sequence and then search for the amino acid sequences with the greatest predicted affinity using quantum annealing. We also use proximal policy optimization to find PBPs with a broader range of physicochemical properties, such as isoelectric point or solubility. Evaluation of the discovered PBPs in MD simulations demonstrates that the peptides have high affinity for two of the plastics: polyethylene and polypropylene. We conclude by describing how our computational approach could be paired with experimental approaches to create a nexus for designing and optimizing peptide-based tools that aid the detection, capture, or biodegradation of MPs. We thus hope that this study will aid in the fight against MP pollution.

ConoDL: a deep learning framework for rapid generation and prediction of conotoxins

Conotoxins, being small disulfide-rich and bioactive peptides, manifest notable pharmacological potential and find extensive applications. However, the exploration of conotoxins' vast molecular space using traditional methods is severely limited, necessitating the urgent need of developing novel approaches. Recently, deep learning (DL)-based methods have advanced to the molecular generation of proteins and peptides. Nevertheless, the limited data and the intricate structure of conotoxins constrain the application of deep learning models in the generation of conotoxins. We propose ConoDL, a framework for the generation and prediction of conotoxins, comprising the end-to-end conotoxin generation model (ConoGen) and the conotoxin prediction model (ConoPred). ConoGen employs transfer learning and a large language model (LLM) to tackle the challenges in conotoxin generation. Meanwhile, ConoPred filters artificial conotoxins generated by ConoGen, narrowing down the scope for subsequent research. A comprehensive evaluation of the peptide properties at both sequence and structure levels indicates that the artificial conotoxins generated by ConoDL exhibit a certain degree of similarity to natural conotoxins. Furthermore, ConoDL has generated artificial conotoxins with novel cysteine scaffolds. Therefore, ConoDL may uncover new cysteine scaffolds and conotoxin molecules, facilitating further exploration of the molecular space of conotoxins and the discovery of pharmacologically active variants.

Screening and identification of antimicrobial peptides from the gut microbiome of cockroach Blattella germanica

BACKGROUND

The overuse of antibiotics has led to lethal multi-antibiotic-resistant microorganisms around the globe, with restricted availability of novel antibiotics. Compared to conventional antibiotics, evolutionarily originated antimicrobial peptides (AMPs) are promising alternatives to address these issues. The gut microbiome of Blattella germanica represents a previously untapped resource of naturally evolving AMPs for developing antimicrobial agents.

RESULTS

Using the in-house designed tool "AMPidentifier," AMP candidates were mined from the gut microbiome of B. germanica, and their activities were validated both in vitro and in vivo. Among filtered candidates, AMP1, derived from the symbiotic microorganism Blattabacterium cuenoti, demonstrated broad-spectrum antibacterial activity, low cytotoxicity towards mammalian cells, and a lack of hemolytic effects. Mechanistic studies revealed that AMP1 rapidly permeates the bacterial cell and accumulates intracellularly, resulting in a gradual and mild depolarization of the cell membrane during the initial incubation period, suggesting minimal direct impact on membrane integrity. Furthermore, observations from fluorescence microscopy and scanning electron microscopy indicated abnormalities in bacterial binary fission and compromised cell structure. These findings led to the hypothesis that AMP1 may inhibit bacterial cell wall synthesis. Furthermore, AMP1 showed potent antibacterial and wound healing effects in mice, with comparable performances of vancomycin.

CONCLUSIONS

This study exemplifies an interdisciplinary approach to screening safe and effective AMPs from natural biological tissues, and our identified AMP 1 holds promising potential for clinical application.

Machine Learning-Driven Methods for Nanobody Affinity Prediction

Because of their high affinity, specificity, and environmental stability, nanobodies (Nbs) have continuously received attention from the field of biological research. However, it is tough work to obtain high-affinity Nbs using experimental methods. In the current study, 12 machine learning algorithms were compared in parallel to explore the potential patterns between Nb-ligand affinity and eight noncovalent interactions. After model comparison and optimization, four optimized models (SVMrB, RotFB, RFB, and C50B) and two stacked models (StackKNN and StackRF) based on nine uncorrelated (correlation coefficient <0.65) optimized models were selected. All the models showed an accuracy of around 0.70 and high specificity. Compared to the other models, RotFB and RFB were not capable of predicting nonaffinitive Nbs with lower precision (<0.44) but showed higher sensitivity at 0.6761 and 0.3521 and good model robustness (F1 score and MCC values). On the contrary, SVMrB, C50B, and StackKNN were able to effectively predict the future nonaffinitive Nbs (specificity >0.92) and reduce the number of true affinitive Nbs (precision >0.5). On the other hand, StackRF showed intermediate model performance. Furthermore, an in-depth feature analysis indicated that hydrogen bonding and aromatic-associated interactions were the key noncovalent interactions in determining Nb-ligand binding affinity. In summary, the current study provides, for the first time, a tool that can effectively predict whether there is an affinity between nanobodies and their intended ligands and explores the key factors that influence their affinity, which could improve the screening and design process of Nbs and accelerate the development of Nb drugs and applications.

An explainable few-shot learning model for the directed evolution of antimicrobial peptides

Due to the persistent threat of antibiotic resistance posed by Gram-negative pathogens, the discovery of new antimicrobial agents is of critical importance. In this study, we employed deep learning-guided directed evolution to explore the chemical space of antimicrobial peptides (AMPs), which present promising alternatives to traditional small-molecule antibiotics. Utilizing a fine-tuned protein language model tailored for small dataset learning, we achieved structural modifications of the lipopolysaccharide-binding domain (LBD) derived from Marsupenaeus japonicus, a prawn species of considerable value in aquaculture and commercial fisheries. The engineered LBDs demonstrated exceptional activity against a range of Gram-negative pathogens. Drawing inspiration from evolutionary principles, we elucidated the bactericidal mechanism through molecular dynamics simulations and mapped the directed evolution pathways using a ladderpath framework. This work highlights the efficacy of explainable few-shot learning in the rational design of AMPs through directed evolution.

Transforming drug discovery: the impact of AI and molecular simulation on R&D efficiency

The process of developing new drugs in the pharmaceutical industry is both time-consuming and costly, making efficiency crucial. Recent advances in hardware and computational methods have led to the widespread application of computational science approaches in drug discovery. These approaches, including artificial intelligence and molecular simulations, span from target identification to pharmacokinetics research, aiming to reduce the likelihood of failure and present lower costs. Machine learning-based methods predict new applications for developing new drugs based on accumulated knowledge, while molecular simulations estimate interactions between drugs and target proteins at the atomic level based on physical laws. Each approach has its advantages and disadvantages, and they complement each other. As a result, the future of computational science approaches in drug discovery is expected to focus on developing new methodologies that integrate these two techniques to enhance the efficiency of drug discovery.

Bioprospecting of culturable marine biofilm bacteria for novel antimicrobial peptides

Antimicrobial peptides (AMPs) have become a viable source of novel antibiotics that are effective against human pathogenic bacteria. In this study, we construct a bank of culturable marine biofilm bacteria constituting 713 strains and their nearly complete genomes and predict AMPs using ribosome profiling and deep learning. Compared with previous approaches, ribosome profiling has improved the identification and validation of small open reading frames (sORFs) for AMP prediction. Among the 80,430 expressed sORFs, 341 are identified as candidate AMPs with high probability. Most potential AMPs have less than 40% similarity in their amino acid sequence compared to those listed in public databases. Furthermore, these AMPs are associated with bacterial groups that are not previously known to produce AMPs. Therefore, our deep learning model has acquired characteristics of unfamiliar AMPs. Chemical synthesis of 60 potential AMP sequences yields 54 compounds with antimicrobial activity, including potent inhibitory effects on various drug-resistant human pathogens. This study extends the range of AMP compounds by investigating marine biofilm microbiomes using a novel approach, accelerating AMP discovery.

Antimicrobial peptides and proteins against drug-resistant pathogens

The rise of drug-resistant pathogens, driven by the misuse and overuse of antibiotics, has created a formidable challenge for global public health. Antimicrobial peptides and proteins have garnered considerable attention as promising candidates for novel antimicrobial agents. These bioactive molecules, whether derived from natural sources, designed synthetically, or predicted using artificial intelligence, can induce lethal effects on pathogens by targeting key microbial structures or functional components, such as cell membranes, cell walls, biofilms, and intracellular components. Additionally, they may enhance overall immune defenses by modulating innate or adaptive immune responses in the host. Of course, development of antimicrobial peptides and proteins also face some limitations, including high toxicity, lack of selectivity, insufficient stability, and potential immunogenicity. Despite these challenges, they remain a valuable resource in the fight against drug-resistant pathogens. Future research should focus on overcoming these limitations to fully realize the therapeutic potential of antimicrobial peptides in the infection control.

ClickGen: Directed exploration of synthesizable chemical space via modular reactions and reinforcement learning

Despite the significant potential of generative models, low synthesizability of many generated molecules limits their real-world applications. In response to this issue, we develop ClickGen, a deep learning model that utilizes modular reactions like click chemistry to assemble molecules and incorporates reinforcement learning along with inpainting technique to ensure that the proposed molecules display high diversity, novelty and strong binding tendency. ClickGen demonstrates superior performance over the other reaction-based generative models in terms of novelty, synthesizability, and docking conformation similarity for existing binders targeting the three proteins. We then proceeded to conduct wet-lab validation on the ClickGen's proposed molecules for poly adenosine diphosphate-ribose polymerase 1. Due to the guaranteed high synthesizability and model-generated synthetic routes for reference, we successfully produced and tested the bioactivity of these novel compounds in just 20 days, much faster than typically expected time frame when handling sufficiently novel molecules. In bioactivity assays, two lead compounds demonstrated superior anti-proliferative efficacy against cancer cell lines, low toxicity, and nanomolar-level inhibitory activity to PARP1. We demonstrate that ClickGen and related models may represent a new paradigm in molecular generation, bringing AI-driven, automated experimentation and closed-loop molecular design closer to realization.

TG-CDDPM: text-guided antimicrobial peptides generation based on conditional denoising diffusion probabilistic model

Antimicrobial peptides (AMPs) have emerged as a promising substitution to antibiotics thanks to their boarder range of activities, less likelihood of drug resistance, and low toxicity. Traditional biochemical methods for AMP discovery are costly and inefficient. Deep generative models, including the long-short term memory model, variational autoencoder model, and generative adversarial model, have been widely introduced to expedite AMP discovery. However, these models tend to suffer from the lack of diversity in generating AMPs. The denoising diffusion probabilistic model serves as a good candidate for solving this issue. We proposed a three-stage Text-Guided Conditional Denoising Diffusion Probabilistic Model (TG-CDDPM) to generate novel and homologous AMPs. In the first two stages, contrastive learning and inferring models are crafted to create better conditions for guiding AMP generation, respectively. In the last stage, a pre-trained conditional denoising diffusion probabilistic model is leveraged to enrich the peptide knowledge and fine-tuned to learn feature representation in downstream. TG-CDDPM was compared to the state-of-the-art generative models for AMP generation, and it demonstrated competitive or better performance with the assistance of text description as supervised information. The membrane penetration capabilities of the identified candidate AMPs by TG-CDDPM were also validated through molecular weight dynamics experiments.

Directed evolution of antimicrobial peptides using multi-objective zeroth-order optimization

Antimicrobial peptides (AMPs) emerge as a type of promising therapeutic compounds that exhibit broad spectrum antimicrobial activity with high specificity and good tolerability. Natural AMPs usually need further rational design for improving antimicrobial activity and decreasing toxicity to human cells. Although several algorithms have been developed to optimize AMPs with desired properties, they explored the variations of AMPs in a discrete amino acid sequence space, usually suffering from low efficiency, lack diversity, and local optimum. In this work, we propose a novel directed evolution method, named PepZOO, for optimizing multi-properties of AMPs in a continuous representation space guided by multi-objective zeroth-order optimization. PepZOO projects AMPs from a discrete amino acid sequence space into continuous latent representation space by a variational autoencoder. Subsequently, the latent embeddings of prototype AMPs are taken as start points and iteratively updated according to the guidance of multi-objective zeroth-order optimization. Experimental results demonstrate PepZOO outperforms state-of-the-art methods on improving the multi-properties in terms of antimicrobial function, activity, toxicity, and binding affinity to the targets. Molecular docking and molecular dynamics simulations are further employed to validate the effectiveness of our method. Moreover, PepZOO can reveal important motifs which are required to maintain a particular property during the evolution by aligning the evolutionary sequences. PepZOO provides a novel research paradigm that optimizes AMPs by exploring property change instead of exploring sequence mutations, accelerating the discovery of potential therapeutic peptides.