Image_3.jpg

Antimicrobial peptides (AMPs) have been recognized for their ability to target processes important for biofilm formation. Given the vast array of AMPs, identifying potential candidates remains a significant challenge, and prompts the need for preliminary in silico investigations prior to extensive in vitro and in vivo studies. We have developed Biofilm-AMP (B-AMP), a curated 3D structural and functional repository of AMPs relevant to biofilm studies. In its current version, B-AMP contains predicted 3D structural models of 5544 AMPs (from the DRAMP database) developed using a suite of molecular modeling tools. The repository supports a user-friendly search, using source, name, DRAMP ID, and PepID (unique to B-AMP). Further, AMPs are annotated to existing biofilm literature, consisting of a vast library of over 10,000 articles, enhancing the functional capabilities of B-AMP. To provide an example of the usability of B-AMP, we use the sortase C biofilm target of the emerging pathogen Corynebacterium striatum as a case study. For this, 100 structural AMP models from B-AMP were subject to in silico protein-peptide molecular docking against the catalytic site residues of the C. striatum sortase C protein. Based on docking scores and interacting residues, we suggest a preference scale using which candidate AMPs could be taken up for further in silico, in vitro and in vivo testing. The 3D protein-peptide interaction models and preference scale are available in B-AMP. B-AMP is a comprehensive structural and functional repository of AMPs, and will serve as a starting point for future studies exploring AMPs for biofilm studies. B-AMP is freely available to the community at https://b-amp.karishmakaushiklab.com and will be regularly updated with AMP structures, interaction models with potential biofilm targets, and annotations to biofilm literature.

The escalating issue of antibiotic resistance has created an urgent global demand within the biomedical field for the discovery of novel antimicrobial molecules as alternatives to traditional antibiotics. Previous studies have reported the identification of potential candidate antimicrobial peptides (AMPs) from extensive bacterial genomes by using deep learning. However, protists, as unique microorganisms capable of thriving in complex ecological environments, represent a largely untapped reservoir of unknown AMP resources that are awaiting systematic exploration. In this study, we harnessed deep learning techniques to identify novel antimicrobial peptides from large-scale protist genomes. Our results indicate that from 2120 protist genome data sets, through a multistage screening process involving sequence redundancy, the application of existing models (C_AMPs_Ptrdict, AMPEP, and AMPidentifier), and the integration of our optimized BERT and CNN models, we ultimately identified 3133 novel candidate AMPs from approximately 6.6 billion sequences. Experimental validation of 18 synthesized candidate peptides demonstrated inhibitory activity against at least one tested bacterial species. This high validation rate, where all 18 tested peptides exhibited inhibitory activity against at least one bacterial species, underscores the exceptional accuracy of the multimodel comprehensive identification approach for obtaining AMPs with antibacterial capability. Among the synthesized peptides, AMP_N2 and AMP_N3 exhibited strong antimicrobial activity in liquid culture while maintaining hemolysis rates below 3%. Our findings indicate that protists are a significant source of novel antimicrobial peptides and that deep learning techniques can be efficiently employed for AMP discovery. To our knowledge, this study represents the first large-scale exploration of novel antimicrobial peptides from over 2000 protist genomes.

The human gut microbiome encodes a large variety of antimicrobial peptides (AMPs), but the short lengths of AMPs pose a challenge for computational prediction. Here we combined multiple natural language processing neural network models, including LSTM, Attention and BERT, to form a unified pipeline for candidate AMP identification from human gut microbiome data. Of 2,349 sequences identified as candidate AMPs, 216 were chemically synthesized, with 181 showing antimicrobial activity (a positive rate of >83%). Most of these peptides have less than 40% sequence homology to AMPs in the training set. Further characterization of the 11 most potent AMPs showed high efficacy against antibiotic-resistant, Gram-negative pathogens and demonstrated significant efficacy in lowering bacterial load by more than tenfold against a mouse model of bacterial lung infection. Our study showcases the potential of machine learning approaches for mining functional peptides from metagenome data and accelerating the discovery of promising AMP candidate molecules for in-depth investigations.

Antimicrobial resistance poses a major threat to public health, prompting the development of alternative therapies such as antimicrobial peptides (AMPs). Protein language models (PLMs) have advanced protein structure and function predictions, facilitating AMP discovery. We developed antimicrobial peptide structural evolution miner (AMP-SEMiner), an AI-driven framework that integrates PLMs, structural clustering, and evolutionary analysis to systematically identify AMPs encoded by small open reading frames and AMP-containing proteins in metagenome-assembled genomes. AMP-SEMiner identified over 1.6 million AMP candidates across diverse environments. Experimental validation showed antimicrobial activity in 9 of the 20 tested candidates, with 5 surpassing antibiotic effectiveness; variant peptides derived from these candidates similarly demonstrated strong antimicrobial efficacy. AMPs from human gut microbiomes revealed both conserved and adaptive evolutionary strategies, reflecting their dynamic ecological roles. AMP-SEMiner thus represents a valuable tool for expanding AMP discovery and has significant potential to inform the development of alternative antimicrobial treatments.

One of the challenges of modern biotechnology is to find new routes to mitigate the resistance to conventional antibiotics. Antimicrobial peptides (AMPs) are an alternative type of biomolecules, naturally present in a wide variety of organisms, with the capacity to overcome the current microorganism resistance threat. Here, we reviewed our recent efforts to develop a new library of non-rationally produced AMPs that relies on bacterial genome inherent diversity and compared it with rationally designed libraries. Our approach is based on a four-stage workflow process that incorporates the interplay of recent developments in four major emerging technologies: artificial intelligence, molecular dynamics, surface-display in microorganisms, and microfluidics. Implementing this framework is challenging because to obtain reliable results, the in silico algorithms to search for candidate AMPs need to overcome issues of the state-of-the-art approaches that limit the possibilities for multi-space data distribution analyses in extremely large databases. We expect to tackle this challenge by using a recently developed classification algorithm based on deep learning models that rely on convolutional layers and gated recurrent units. This will be complemented by carefully tailored molecular dynamics simulations to elucidate specific interactions with lipid bilayers. Candidate AMPs will be recombinantly-expressed on the surface of microorganisms for further screening via different droplet-based microfluidic-based strategies to identify AMPs with the desired lytic abilities. We believe that the proposed approach opens opportunities for searching and screening bioactive peptides for other applications.

Drosophila melanogaster has been a valuable model for dissecting the molecular architecture of innate immunity. However, the family Drosophilidae encompasses over 4000 species, spanning deep evolutionary divergences and diverse ecologies. Here, we use immune challenge with the Gram-negative pathogen Providencia rettgeri to investigate the conservation and evolution of immune responses in three non-model drosophilid species that diverged from D. melanogaster over 45 million years ago-Hirtodrosophila cameraria, H. confusa, and Scaptodrosophila deflexa. We find that all three species retain a core set of immune signaling and recognition genes, but exhibit substantial variation in effector gene content and inducibility. In particular, Scaptodrosophila deflexa lacks orthologs of multiple antimicrobial peptides (AMPs) known from D. melanogaster, including DptA, AttA, and AttC, and shows little transcriptional response to bacterial challenge with Providencia rettgeri. In contrast, both of the Hirtodrosophila species exhibit substantial transcriptional responses, including strong induction of canonical Imd pathway genes. Microbiome profiling of our samples revealed higher Providencia abundance in H. cameraria, and high levels of the defensive symbiont Spiroplasma in S. deflexa-potentially explaining differences in infection outcome. Our combined annotation and expression analysis of these species also allowed us to identify 20 novel AMP-like candidates, many with structural features like known AMPs. Our study demonstrates the feasibility of functional immune analyses in non-model Drosophila species and reveals striking lineage-specific differences in immune gene repertoire and expression. These findings highlight the importance of non-model, wild-derived samples for uncovering novel immune effectors and understanding evolutionary forces shaping insect immunity.

All living organisms, ranging from microorganisms to plants and mammals, have evolved mechanisms to actively defend themselves against pathogen attack. A wide range of biological activities have been attributed to plant antimicrobial peptides (AMPs) including growth inhibitory effects on a broad range of fungi, Gram-positive and Gram-negative bacteria, viruses, neoplasic cells and parasitic protozoa. Classes of AMPs, their mechanisms of action, biological activity, and cytotoxicity towards host cells are discussed. A particular focus regards AMP candidates with potential for use in defense against biological warfare agents. This field is young, but provides additional stimulus to consideration of these molecules as a new class of therapeutic agents and promises to revolutionize treatment of many infectious diseases.

Machine learning and molecular simulation ascertain antimicrobial peptide against Klebsiella pneumoniae from public database

The Smart House Planner is an AI-powered solution designed to simplify the process of residential home design by combining intelligent planning, 3D modelling, and real-time cost estimation through an interactive conversational interface. With the rising demand for personalized housing solutions and the complexity of balancing aesthetics, functionality, and budget, there is a growing need for systems that can automate architectural planning while remaining user-friendly. This project addresses that challenge by integrating Gemini AI for intelligent design recommendations, Blender for automated 3D visualization, and a cost estimation engine for transparent financial planning. The system enables users to specify requirements such as housing type (1BHK, 2BHK, 3BHK, Duplex, Villa), dimensions, and style preferences via a modern Django-based web interface. Gemini AI processes these inputs to generate optimized layouts by considering space utilization, and functional flow. These layouts are then transformed into realistic, interactive 3D models using Blender, allowing users to explore their prospective homes virtually. Simultaneously, the cost estimation module calculates material, labour, costs, producing downloadable reports that support informed decision-making and budget control. By unifying AI-driven design intelligence, real-time 3D modelling, and detailed cost breakdowns, Smart House Planner provides a comprehensive platform that enhances user engagement, reduces reliance on professional intervention, and democratizes access to architectural planning. This solution demonstrates how AI, visualization technologies, and web-based interfaces can transform the way individuals approach housing design, making the process more efficient, transparent, and accessible to both professionals and non-experts.

Antimicrobial peptides (AMPs) are a diverse class of short, often cationic biological molecules that present promising opportunities in the development of new therapeutics to combat antimicrobial resistance. Newly developed in silico methods offer the ability to rapidly discover numerous novel AMPs with a variety of physiochemical properties. Herein, using the rAMPage AMP discovery pipeline, we bioinformatically identified 51 AMP candidates from amphibia and insect RNA-seq data and present their in-depth characterization. The studied AMPs demonstrate activity against a panel of bacterial pathogens and have undetected or low toxicity to red blood cells and human cultured cells. Amino acid sequence analysis revealed that 30 of these bioactive peptides belong to either the Brevinin-1, Brevinin-2, Nigrocin-2, or Apidaecin AMP families. Prediction of three-dimensional structures using ColabFold indicated an association between peptides predicted to adopt a helical structure and broad-spectrum antibacterial activity against the Gram-negative and Gram-positive species tested in our panel. These findings highlight the utility of associating the diverse sequences of novel AMPs with their estimated peptide structures in categorizing AMPs and predicting their antimicrobial activity.

Discovery of naturally inspired antimicrobial peptides using deep learning.

In a previous study, we analyzed the transcriptome of Scolopendra subspinipes mutilans using next-generation sequencing technology and identified several antimicrobial peptide candidates. One of the peptides, scolopendrasin V, was selected based on the physicochemical properties of antimicrobial peptides using a bioinformatics strategy. In this study, we assessed the antimicrobial activities of scolopendrasin V using the radial diffusion assay and colony count assay. We also investigated the mode of action of scolopendrasin V using flow cytometry. We found that scolopendrasin V's mechanism of action involved binding to the surface of microorganisms via a specific interaction with lipopolysaccharides, lipoteichoic acid, and peptidoglycans, which are components of the bacterial membrane. These results provide a basis for developing peptide antibiotics.

Previously, we performed an in silico analysis of the Periplaneta americana transcriptome. Antimicrobial peptide candidates were selected using an in silico antimicrobial peptide prediction method. It was found that periplanetasin-5 had antimicrobial activity against yeast and grampositive and gram-negative bacteria. In the present study, we demonstrated the anti-inflammatory activities of periplanetasin-5 in mouse macrophage Raw264.7 cells. No cytotoxicity was observed at 60 μg/ml periplanetasin-5, and treatment decreased nitric oxide production in Raw264.7 cells exposed to lipopolysaccharide (LPS). In addition, quantitative RT-PCR and enzyme-linked immunosorbent assay revealed that periplanetasin-5 reduced cytokine (tumor necrosis factor-α, interleukin-6) expression levels in the Raw264.7 cells. Periplanetasin-5 controlled inflammation by inhibiting phosphorylation of MAPKs, an inflammatory signaling element, and reducing the degradation of IκB. Through LAL assay, LPS toxicity was found to decrease in a periplanetasin-5 dose-dependent manner. Collectively, these data showed that periplanetasin-5 had antiinflammatory activities, exemplified in LPS-exposed Raw264.7 cells. Thus, we have provided a potentially useful antibacterial peptide candidate with anti-inflammatory activities.

A novel antimicrobial peptide derived from fish goose type lysozyme disrupts the membrane of Salmonella enterica

Structural and mechanistic divergence in LL-37, HNP-1, and Magainin-2: An integrated computational and biophysical analysis