Biology subjects

Islam, T.

Publications and source records attributed to Islam, T..

At least 19 recordsLinked to original sources

bioRxiv2026

A relational architecture of the human genome

Complex systems exhibit organization not only through their components but also via their interdependencies. Local nucleotide dependencies underpin regulatory motifs and RNA structures, yet principles governing a higher-order organization across the human genome remain unknown. Here we systematically perturbed every position of the com- plete telomere-to-telomere human genome using Nucleotide Transformer v3 with an ex- ascale compute cluster, generating nucleotide-resolution dependency maps that reveal a directed, weighted network of position-to-position genomic relationships. This network recapitulates known biological features and uncovers recurrent, directional architec- tures termed 'relons', spanning coding, non-coding, and unannotated regions. These re- lational modules reflect evolutionary and functional constraints and associate with hu- man genetic variation and pathogenic variants. Our findings suggest a higher-order rela- tional layer of genome organization, providing a framework to interpret DNA as a complex system of dependencies beyond linear and 3D structure, with implications for under- standing genetic variation and disease.

genomics↗

bioRxiv2026

Conditional Spatial Classification of Expert-Confirmed Interictal Epileptiform Discharge Epochs: An EEG-ECG Ablation and SHAP Analysis

Interictal epileptiform discharges (IEDs) are diagnostically important EEG abnormalities observed between seizures. This study addresses a conditional spatial-classification task where every analyzed four-second epoch had already been reviewed and confirmed by experts as containing an IED, and the model assigned that epoch to one of five predefined scalp-distribution categories (generalized, frontal, temporal, occipital, or centro-parietal). The analysis therefore does not evaluate IED-versus-non-IED detection. After preprocessing, 2,514 IED-labelled epochs were analyzed using identical stratified epoch-level partitions, SMOTE based training, 26 handcrafted features per included channel, and multiple machine-learning classifiers. A staged channel ablation compared 19-channel scalp EEG, 21-channel EEG with ECG, and the complete 29-channel input containing scalp EEG, referential, ECG, and EMG channels. The best EEG-only result was obtained with linear discriminant analysis (88.89% test accuracy). CatBoost achieved 93.25% on EEG with ECG channel and 94.44% with the whole channel set. All eight directly comparable classifiers showed numerically higher test accuracy after ECG channel was added; for CatBoost, the increase was 6.35 percentage points. In the EEG with ECG channel, CatBoost model on ECG channel on right and left arm received respectively 15.79% and 15.12% of normalized global SHAP attribution, and beta-band power was the leading of all features (18.76%). These SHAP values indicate model-specific predictive contributions and do not establish physiological biomarkers, causal autonomic mechanisms, or clinical localization. The findings support a limited methodological conclusion which is ECG-derived features were associated with improved internal epoch-level categorization of expert-confirmed IED epochs. They do not establish IED detection, artifact rejection, independent EMG effects, or generalization to unseen patients.

bioengineering↗

bioRxiv2026

Residual Multi-Modal Learning for Pan-Breast-Cancer Drug Response Prediction

Predicting drug sensitivity across diverse cancer cell lines remains a fundamental challenge in precision oncology, particularly for data-scarce cell lines where per-cell-line models overfit and lookup-table approaches cannot generalise to unseen biological contexts. We present DL4DR, a Two-Tower Residual Late Fusion deep learning model that addresses this challenge through content-based, identity-free genomic conditioning. The Cell Line Tower encodes each cell line as a 3 x 139 x 139 genomic image - encoding gene expression, mutation severity, and copy-number variation as RGB channels - using a convolutional encoder that maps directly from biological content, never from a cell line ID. The Compound Tower combines three complementary molecular representations: D-MPNN graph message passing, ORNN octave convolutional image features, and an ECFP hard-memorization head that preserves activity-cliff resolution. Predictions are composed as a residual sum: f = fhard +{lambda} (zC) {middle dot} fresidual, where the learned gate{lambda} modulates how much interaction signal supplements the memorization baseline. Evaluated across 51 breast cancer cell lines (136,342 records), Residual Fusion outperforms the ECFP-Only baseline in 48/51 cell lines (94.1%), with {Delta}R2 > 0.02 in 26/51 (51.0%). On the leave-cell-line-out split - the decisive test of genomic generalisation - the mean {Delta}R2 = +0.016 across all 51 lines demonstrates that the genomic encoder learns transferable biological signal beyond cell line identity. External validation on 601 cell lines across 27 cancer tissue types (CellTiter-Glo dataset; 0 cell line overlap with training) achieves median R2 = 0.627, within the range of the internal random-split performance (R2 = 0.61-0.69), confirming pan-cancer generalisation. GradCAM interpretability on the Cell Line Tower recovers TP53 among the top-five cross-cell-line genomic activators (5/51 cell lines) alongside several uncharacterised candidate genes (e.g. FSIP2, 6/51) - without any prior pathway annotation - providing partial biological validation of the learned representation, while also indicating that a substantial share of the encoders top-ranked signal corresponds to genes with no current annotation as breast cancer drivers. Code and data are available at https://github.com/bayjuan5/DL4DR.

bioinformatics↗

bioRxiv2026

First occurrence of Corynespora cassiicola infecting chia plant in Bangladesh and its sensitivity to selected fungicides

Chia (Salvia hispanica L.) is an emerging crop in Bangladesh valued for its medicinal properties and economic significance. In March 2024, target spot-like symptoms were observed in an experimental chia field (24.75{degrees} N, 90.50{degrees} E) at Bangladesh Agricultural University in Mymensingh, Bangladesh with disease incidence ranging from 23% to 47% across approximately 0.25 ha. Initially appearing as brick-red spots, these symptoms developed into target-shaped concentric rings, affecting leaves, stems, and inflorescences. A total of 24 fungal isolates were recovered from infected tissue; two representative isolates (BGECh-3 and BGECh-4) were randomly selected for details characterization. Pathogen identity was established through morphological traits, multilocus phylogenetic analysis of internal transcribed spacer (ITS) and elongation factor 1-alpha (EF-1) genes sequence, and pathogenicity confirmation through Kochs postulates, collectively identifying the causal agent as Corynespora cassiicola. The isolates demonstrated a broad host range, successfully infecting brinjal, chili, bottle gourd, country bean, tomato, and soybean. In vitro fungicide sensitivity assays with seven commercial fungicides showed that both isolates were highly sensitive to Goldzim (50% carbendazim), which completely inhibited mycelial growth at 10 {micro}g mL-{superscript 1}. Conza (10% Hexaconazole) and Amister top (18.2% azoxystrobin + 11.4% difenoconazole) reduced growth by up to 85% and 67%, respectively at equal concentration. Other fungicides showed comparatively lower efficacy even at higher concentrations. This study represents the first report of target spot disease of chia caused by C. cassiicola in Bangladesh and provides insights for effective disease management strategies.

molecular biology↗

bioRxiv2025

Rutin derived microbiota metabolite 3,4 dihydroxybenzoic acid restores antibiotic susceptibility in XDR Gram negative pathogens with a system wide susceptibility signature

Antibiotic potentiators are a practical route to extend the utility of existing drugs against multidrug resistant Gram-negative pathogens, but most natural compound combinations remain mechanistically under defined and are rarely mapped at the level of whole cell susceptibility determinants. Here we apply a metabolite guided potentiator strategy based on gut microbiota derived breakdown products of the dietary polyphenol rutin, and we integrate potentiation phenotyping with system wide genetic susceptibility mapping to guide combination design. Three rutin metabolites, 3,4 dihydroxybenzoic acid (DHBA), 2,4,6 trihydroxybenzoic acid (THBA), and 3,4 dihydroxyphenylacetic acid (DOPAC), were screened against a CDC ARLG reference panel composed of multidrug resistant and extensively drug resistant (XDR) isolates, identifying DHBA as the most broadly active candidate, including against XDR Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Escherichia coli. DHBA restored susceptibility to selected antibiotic classes in resistant Gram-negative pathogens, including an approximately sixteen-fold reduction in colistin MIC in mcr 1 positive E. coli. To better understand the potentiation mechanism, we performed a large-scale genetic susceptibility screen of 316 E. coli single gene knockouts and defined a DHBA response architecture enriched for envelope and transport determinants with additional contributions from central metabolism and information processing pathways. Comparative mapping under colistin exposure revealed a distinct susceptibility architecture with limited overlap, supporting the concept that both potentiators and antibiotics engage secondary cellular systems beyond canonical primary mechanisms. In a Mini Bioreactor Array gut community model, DHBA produced a more conserved community shift than colistin. Finally, DHBA-antibiotic combinations improved outcomes in infection relevant models, including improved survival in Galleria mellonella, reduced intestinal burden in Caenorhabditis elegans, and reduced bacterial burden in an ex vivo porcine burn wound infection model. Collectively, these findings support a systems-based framework for developing mechanistically informed potentiator antibiotic-combinations to extend the lifespan of existing antibiotics.

microbiology↗

bioRxiv2025

Genolator: A Multimodal Large Language Model Fusing Natural Language, Genomic, and Structural Tokens for Protein Function Interpretation

BackgroundDecoding the genetic code to unveil its genome functionality is a monumental task which would greatly advance the understanding of disease mechanisms and development of targeted treatment approaches. Although large language models (LLMs) have transformed natural language processing across diverse domains, translating the complex language of DNA into human-readable form remains challenging due to genomic data complexity and unexplored regions of the human genome. Current (genomic) language models either have a solid understanding of natural language or of the genomic code. Models fusing both aspects are largely lacking. ResultsHere we present Genolator, a multimodal large language model that integrates embeddings from DNA sequences, amino acid sequences, and protein structures with natural language queries. Fine-tuned on over 370,000 question-answer pairs generated using abstracted Gene-Ontology (GO) terms, Genolator effectively answers queries regarding protein subcellular localization, molecular function, and biological processes. Evaluation demonstrates high accuracy in confirming or denying protein function associations, outperforming baseline models such as openly available allrounder LLMs like GPT 4.1 as well as smaller domain specific models integrating knowledge from foundation models like Evo2 and ESM-2. Explorations of the Genolators hidden states unveil a biologically and linguistically plausible organization of its learned representations. ConclusionGenolator enhances accessibility to genomic information by enabling natural language interaction with protein data, facilitating biological discovery, and clinical research. It represents a step towards bridging genomic code and human language through the integration of a multimodal LLM.

bioinformatics↗

bioRxiv2025

αTAT1 defines a microtubule mechanosensing axis that drives fibroblast durotaxis and fibrosis across organs

Durotaxis, the directed migration of cells along gradients of extracellular stiffness, drives tissue fibrosis by recruiting fibroblasts to stiffened injury sites where hey differentiate into myofibroblasts and deposit scar tissue. While actin cytoskeletal tension and focal adhesion dynamics have been implicated in this process, the contribution of microtubules to cellular mechanosensing and durotaxis has remained undefined. Here, we uncover TAT1-mediated microtubule acetylation as a master regulator of fibroblast mechanosensing and stiffness-directed durotaxis. By catalyzing -tubulin K40 acetylation, TAT1 confers the structural flexibility required for directional microtubule alignment and persistent polarity along stiffness gradients, enabling fibroblasts to sense mechanical cues and initiate profibrotic programs. Loss of TAT1 abolishes K40 acetylation, disrupts focal adhesion FAK signaling, and suppresses YAP nuclear localization, thereby uncoupling extracellular matrix stiffness from downstream mechanotransduction. Global or fibroblast-specific deletion of TAT1 markedly reduces fibroblast durotaxis and myofibroblast accumulation, and protects mice from lung, dermal, and kidney fibrosis in experimental models, without affecting inflammation or vascular integrity. Together, our findings define TAT1-dependent microtubule mechanosensing as a central cytoskeletal pathway coupling fibroblast mechanobiology to organ fibrosis in vivo, positioning the TAT1 catalytic domain as a novel mechano-therapeutic target in fibrotic disease.

cell biology↗

bioRxiv2025

Genomic insights into multidrug resistant Escherichia coli from bovine mastitis in Bangladesh

BackgroundMastitis poses a significant threat to dairy industry and public health due to the emergence of multidrug-resistant (MDR) Escherichia coli. This study provides a genomic characterization of two MDR E. coli strains, MBBL4 and MBBL5, from bovine mastitis in Bangladesh, highlighting their evolutionary relationships, resistome, and virulome. MethodsSpecies-level identification of MBBL4 and MBBL5 was confirmed using biochemical assays, VITEK-2 system, and 16S rRNA gene sequencing. Antimicrobial susceptibility profiling was conducted to determine their resistance patterns. Whole genome sequencing (WGS) and comprehensive genomic analysis were performed for phylogenetic, comparative genomics, mobile genetic elements (MGEs), antimicrobial resistance genes (ARGs), and virulence factor genes (VFGs) analyses. ResultsBoth isolates exhibited extensive MDR patterns, showing resistance to ten antibiotics. Phylogenetic and ANI analyses showed that MBBL4 clustered with mastitis-associated and human bacteremia strains of E. coli, while MBBL5 was closely related to wildlife-associated strains, reflecting divergent evolutionary lineages. Pangenome analysis revealed an open pangenome structure, indicating high genetic diversity, with MBBL4 harboring 21 unique genes and MBBL5 possessing nine unique genes. Both genomes harbored numerous ARGs spanning over 11 antibiotic classes, and VFGs, predominantly associated with adherence and secretion systems, underscoring their extensive resistome, virulome, and adaptive potentials. Abundant MGEs (plasmids, prophages, insertion sequence elements and genomic islands) further underscored the role of horizontal gene transfer in driving resistance and virulence in these strains. ConclusionThis study highlights the zoonotic potential and adaptive capacity of MDR E. coli from bovine mastitis in Bangladesh driven by resistome, virulome, and mobile genetic elements. These findings highlight the urgent need for One Health-based genomic surveillance to mitigate MDR E. coli transmission from dairy farms to humans and the environment.

genomics↗

bioRxiv2025

Systematic Optimization Enables Near-Perfect In Vitro Transformation Efficiencies for Spirodela polyrhiza (Greater Duckweed)

The in vitro transformation of plants, or the delivery of foreign genetic material that is incorporated into their genomes, represents a powerful tool both for elucidating genotype-phenotype relationships and for generating plant cultivars which have desirable traits for agriculture and/or biotechnological applications. However, outside of a few model species, the processes involved in transformation are often inefficient and can take months to perform for many plant species, with several bottlenecks occurring at the different stages of calli induction, genetic transfection, and plant regeneration. While duckweeds - aquatic monocots whose species include some of the smallest and fastest-growing flowering plants on the planet - have distinguished themselves with several emerging biotechnological applications, they too are the subject of conflicting reports regarding their transformation potential and ability to be genetically manipulated. Here, we synthesized and optimized the protocols for in vitro transformation of duckweed Spirodela polyrhiza (Greater Duckweed) from start-to-finish: achieving >90% - 100% efficiencies for each of calli induction; transient and stable genetic transformation; visual marker-free selection of transformants; and regeneration of genetically modified plants with stable transgene expression for over 100 generations - and which in S. polyrhiza can be achieved over the course of weeks instead of months. The integrated, streamlined approaches for all stages of in vitro transformation overcome many bottlenecks and can help to pave the way for high-throughput functional genomics studies and synthetic biology applications in this biotechnologically-important species.

plant biology↗

bioRxiv2025

Bacterial species and surface structures shape gene transfer and the transcriptional landscape during early conjugation

Conjugative plasmids drive bacterial evolution and niche adaptation, yet how their active acquisition reshapes host transcription remains poorly understood. Most studies focus on stable plasmid carriage, overlooking the dynamic transcriptional changes during conjugation itself. Here, we show that active RP4 conjugation triggers an immediate, host- and surface factor-dependent transcriptional response. This includes activation of non-SOS stress pathways, motility, exopolysaccharide production, anaerobic respiration, and metabolic adaptation. These responses do not inhibit conjugation, suggesting they serve to maintain host homeostasis. Capsule expression blocks these responses by preventing conjugation, and also inhibits RP4 activation by physically blocking donor-recipient contact. Unexpectedly, RP4 overcomes this barrier by exploiting single-cell variation in recipient capsule thickness, successfully conjugating with thin-capsulated recipients. These findings reveal a striking interplay between plasmid, host, and surface architecture in shaping the conjugation transcriptional landscape, with broad implications for plasmid dissemination and bacterial evolution.

microbiology↗

bioRxiv2025

Schistosome exposure and diet induced effects on candidate immune gene expression in an African snail vector

Schistosoma mansoni is a parasitic helminth that is vectored through freshwater snails. While the anti-schistosome defense of the South American snail, Biomphalaria glabrata, is well studied, little is known about the immune response of the African snail, Biomphalaria sudanica. We measured expression of five candidate immune genes in B. sudanica 8, 24, and 72 hours post-exposure to S. mansoni using reverse transcription quantitative PCR. Expression patterns of resistant snails were compared to susceptible snails and those sham exposed. We also assessed how diet (lettuce vs. pellet) affected expression of three genes, given prior findings that pellet-fed snails were more susceptible to S. mansoni. Results indicated that resistant snails constitutively expressed higher levels of superoxide dismutase 1 (SOD1) than susceptible snails, consistent with expression patterns of resistant B. glabrata. Parasite-induced expression occurred at 8 hours in SOD1, biomphalysin, thioester protein 1 (TEP1), and granulin (GRN); however, for biomphalysin and TEP1, induced expression was only detected for susceptible snails. At 24 hours, biomphalysin expression increased in exposed resistant snails, and at 72 hours, all exposed snails decreased biomphalysin expression compared to controls. Parasite-induced expression of SOD1, biomphalysin, TEP1, and GRN supports the hypothesis that these genes play a role in B. sudanica anti-schistosome defense, however increased expression does not necessarily yield clearance of S. mansoni. SOD1 expression was higher in lettuce-fed snails at 8 and 24 hours, consistent with their greater resistance. Together, these results demonstrate the conserved and unique aspects of the B. sudanica anti-schistosome response.

microbiology↗

bioRxiv2025

Plant silicon defence disrupts cryptic colouration in an insect herbivore by restricting carotenoid sequestration into the haemolymph

O_LICryptic colouration is a primary anti-predation strategy in herbivorous insects. Achieving crypsis often requires acquiring dietary carotenoids--tetraterpene pigments vital for plant colouration and photoprotection. Silicon (Si) accumulation in plants makes tissues tougher and less digestible for insects, but its effect on plant pigment levels remains unclear. It is also unknown whether feeding on silicified plants impairs insects ability to sequester carotenoids and achieve crypsis. C_LIO_LIUsing the model grass Brachypodium distachyon, we demonstrate that the cotton bollworm (Helicoverpa armigera) larvae exhibited stunted growth and reduced carotenoid sequestration, particularly lutein, into their hemolymph when feeding on Si- supplemented (+Si) plants. This reduction led to distinct body-colour morphs: larvae feeding on +Si plants developed brown colouration, contrasting sharply with the green leaves, whereas larvae feeding on Si-free (-Si) plants exhibited green cryptic colouration that blended seamlessly with the foliage. C_LIO_LIPlant leaves contained various carotenoids (neoxanthin, violaxanthin, {beta}-carotene and lutein) and chlorophylls (a and b), but larvae only sequestered {beta}-carotene and lutein into their haemolymph while excreting substantial amounts of pigments, regardless of plant Si status. Under insect-free conditions, +Si plants had lower carotenoid and chlorophyll contents than -Si plants. However, following insect herbivory, pigment levels in -Si and +Si plants equalised. Insect herbivory also increased Si accumulation in +Si plants. C_LIO_LIOur findings provide novel evidence that plant Si defences can disrupt cryptic colouration in insect herbivores by restricting carotenoid sequestration from host plant tissues. This disruption could increase insect visibility to predators, potentially elevating their risk of predation. C_LI

ecology↗

bioRxiv2025

Computational investigation unveils pathogenic LIG3 non-synonymous mutations and therapeutic targets in acute myeloid leukemia

Single nucleotide polymorphisms (SNPs) in DNA repair genes can impair protein structure and function, contributing to disease development, including cancer. Non-synonymous SNPs (nsSNPs) in the LIG3 gene are linked to genomic instability and increased cancer risk, particularly acute myeloid leukemia (AML). This study aims to identify the most deleterious nsSNPs in the LIG3 gene and potential therapeutic targets for DNA repair restoration in AML. We employed in-silico computational methods to analyze LIG3 nsSNPs, using PredictSNP and Mutation3D to assess pathogenicity. Subsequently, molecular docking and dynamics simulations were conducted to evaluate ligand-binding affinities and protein stability. Out of the 12,191 mapped SNPs, 132 were nsSNPs located in the coding region. Among these, 18 nsSNPs were identified as detrimental including 12 destabilizing and 6 stabilizing nsSNPs. Nine cancer-associated nsSNPs, including L381R and R528C, were predicted due to their structural and functional impacts. Further analysis revealed key phosphorylation and methylation sites, such as 529S and 224R. Molecular dynamics simulations highlighted stable interactions of compounds AHP-MPC and DM-BFC with wild-type and R528C mutant LIG3 proteins, while R671G and V781M mutants showed instability. Protein-protein interaction networks and functional enrichment linked LIG3 to DNA repair pathways. Kaplan-Meier analysis associated high LIG3 expression with improved survival in breast cancer and AML, suggesting its role as a prognostic biomarker. This study emphasizes the mutation-specific effects of LIG3 nsSNPs on protein stability and ligand interactions. We recommend identifying DM-BFC to advance personalized medicine approaches for targeting deleterious variants, following in vitro and in vivo validation for AML treatment.

cancer biology↗

bioRxiv2025

Enterococcus faecium MBBL3 Exhibits Promising Probiotic Potential and Antimicrobial Efficacy Against Bovine Mastitis-Associated Escherichia coli and Klebsiella pneumoniae

Enterococcus faecium, a promising probiotic, combats pathogens, supports gut health, strengthens immunity, and provides a natural approach to address the escalating global challenge of antimicrobial resistance. This study aimed to investigate the genome of E. faecium MBBL3, isolated from healthy cow milk, to assess its probiotic potential and antimicrobial activity against pathogens causing bovine mastitis. The strain was analyzed through whole genome sequencing, along with in-vitro and in-silico assessments were conducted to determine its antimicrobial efficacy against bovine mastitis pathogens, Klebsiella pneumoniae MBBL2 (Kp MBBL2) and Escherichia coli MBBL4 (Ec MBBL4). The genome assembly and functional annotations uncovered many important probiotic traits in MBBL3, where genome comparison revealed its high genetic similarity with other Enterococcus strains. MBBL3 demonstrated the ability to ferment a wide range of carbohydrates and possessed 76 carbohydrate-active enzyme-related genes, including five key CAZy families namely GH73, GH18, CBM50, CE4, and AA10. It also possessed importance genes for bile salt and acid tolerance, stress resistance, and surface adhesion. Additionally, MBBL3 contained metabolite regions involved in the biosynthesis of antimicrobial compounds such as 2,4-DAPG, aborycin, enterocin NKR-5-3B, and sodorifen, and bacteriocin gene clusters for sactipeptides, Enterolysin_A, and UviB. Safety assessments indicated low pathogenic potential, while in-vitro assays demonstrated antibiotic susceptibility and suppressed the growth of Kp MBBL2 and Ec MBBL4, respectively. Its bacteriocin compound Enterolysin_A exhibited strong molecular interactions with virulence proteins of these mastitis pathogens. Therefore, the promising probiotic potential and antimicrobial efficacy of E. faecium MBBL3, especially against mastitis pathogens combined with its safety, position it as a valuable candidate for therapeutic applications. Key pointsO_LIE. faecium MBBL3 genosme showed high similarity with other species of this genera. C_LIO_LIGenetic makeup of MBBL3 revealed its ability to survival and adaptation in different niches including hosts gut. C_LIO_LIIn-vitro and in-silico study results, along with several genes linked to antimicrobials demonstrated its ability to combat against mastitis pathogens. C_LI

microbiology↗

bioRxiv2024

ConvCGP: A Convolutional Neural Network to Predict Genotypic Values of Rice Traits from Compressed Genome-Wide Polymorphisms

The growing size of genome-wide polymorphism data in animal and plant breeding has raised concerns regarding computational load and time, particularly when predicting genotypic values for target traits using genomic prediction. Although several deep learning and conventional methods, including dimensionality reduction techniques, such as principal component analysis (PCA) and autoencoders, have been proposed to address these challenges by selecting subsets of polymorphisms or compressing high-dimensional data for predictive analysis. However, these methods are often computationally intensive and time-consuming. A major challenge in applying high-dimensional genomic data directly to deep-learning models is the substantial computational cost and time required for hyperparameter tuning and model training. To address these limitations, we propose a novel deep learning approach that combines convolutional neural networks (CNNs) to predict the genotypic data of target traits with autoencoders to compress high-dimensional genome-wide polymorphism data. We tested this framework on high-dimensional rice datasets, focusing on agronomic trait prediction. By combining CNNs with autoencoders, our framework outperformed other machine-learning methods and recently proposed compression methods, demonstrating its potential to efficiently address the computational challenges associated with high-dimensional genomic data.

bioinformatics↗

bioRxiv2024

High-Throughput Screening Reveals Potential Inhibitors Targeting Trimethoprim-Resistant DfrA1 Protein in Klebsiella pneumoniae and Escherichia coli

The DfrA1 protein provides trimethoprim resistance in bacteria, especially Klebsiella pneumoniae and Escherichia coli, by modifying dihydrofolate reductase, which reduces the binding efficacy of the antibiotic. Thus, this study aimed to identify inhibitors of the trimethoprim-resistant DfrA1 protein through high-throughput computational screening of 3,601 newly synthesized chemical compounds sourced from the ChemDiv database. We conducted high-throughput computational optimization and screening of a library containing 3,601 compounds against the DfrA1 protein from K. pneumoniae and E. coli to identify potential drug candidates (DCs). Through this extensive approach, we identified six promising DCs, labeled DC1 to DC6, as potential inhibitors of DfrA1. Each DC demonstrated strong initial binding affinity and favorable chemical interactions with the DfrA1 binding sites when compared to the effective drug Iclaprim (effective antibiotic against DfrA1), used as a control. To validate these findings, we further investigated the molecular mechanisms of inhibition, focusing on the thermodynamic properties of the promising DCs. Furthermore, molecular dynamics simulation (MDS) validated the inhibitory efficacy of these six DCs against the DfrA1 protein. Our results showed that DC4 (an organoflourinated compound) and DC6 (a benzimidazol compound) showed superior efficacy against the DfrA1 protein than the control drug, particularly regarding stability, solvent-accessible surface area, solvent exposure, polarity, and binding site interactions, which influence their residence time and efficacy. Overall, findings of this study suggest that DC4 and DC6 have the potential to act as inhibitors against the DfrA1, offering promising prospects for the treatment and management of infections caused by trimethoprim-resistant K. pneumoniae and E. coli in both humans and animals.

bioinformatics↗

bioRxiv2024

Natural Bacteriocins as Potential Drug Candidates Targeting Core Proteins in Mastitis Pathogens of Dairy Cattle

Mastitis poses a major challenge in the dairy industry, with rising antibiotic-resistant strains underscoring the urgent need for alternative antimicrobial strategies. This study aimed to (i) identify essential core proteins in clinical mastitis (CM)-causing pathogens using genomic approach, and (ii) assess the efficacy of natural antimicrobial peptides as novel therapeutic agents targeting the selected core proteins for the rational management of mastitis in dairy cows. Through a core genomic analysis of 16 CM-causing pathogens, including strains of Staphylococcus aureus, S. warneri, Streptococcus agalactiae, S. uberis, Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, P. putida, and P. asiatica, we identified 65 core proteins shared among these pathogens. Among them, ten proteins including PhoH, TrpB, FtsZ, HslV, HupB, RibH, InfA, MurA, GlxK, and Rho were found to be essential for the survival and virulence of these pathogens. Importantly, further novelty, resistance, and virulence assessments identified Rho and HupB as potential therapeutic targets. A comprehensive screening of 70 bacteriocin peptides (BPs) revealed 14 BPs that effectively interacted with both Rho and HupB proteins. Further analysis showed that BP8 and BP32 disrupt Rho protein function by blocking transcription termination process, while BP8, BP39, and BP40 prevent HupB from binding to DNA. These findings confirm the promising stability and efficacy of BP8 against both target proteins in CM-pathogens, highlighting it as a promising broad-spectrum therapeutic agent. Our computational study identified Rho and HupB as key proteins in CM-causing pathogens, which can be targeted by natural bacteriocins like BP8, suggesting its potential for developing effective and sustainable therapeutics against mastitis in dairy cattle. Author SummaryMastitis poses a significant threat to the global dairy industry, with rising antibiotic resistance necessitating alternative therapeutic strategies. This study identified essential core proteins in clinical mastitis-causing pathogens through a genomic approach and evaluated natural antimicrobial peptides (bacteriocins) as novel therapeutic agents. Through a core-genomic analysis, Rho and HupB were identified as key therapeutic targets. Bacteriocin peptides such as BP8 demonstrated promising efficacy by disrupting regular transcription termination process and DNA replication, offering a promising solution for next-generation mastitis therapies. The findings underscore the potential of BP8 as a sustainable, broad-spectrum antimicrobial agent, contributing to the rational management of mastitis in dairy cattle.

bioinformatics↗

bioRxiv2024

Androgen receptor modulation of vocal circuitry in Alston's singing mouse

Animal courtship and aggressive displays are dramatic, often sexually dimorphic behaviors that require the coordinated modulation of diverse motivational and motor circuits. In Alstons singing mice, a novel and elaborate advertisement vocalization is sexually dimorphic and steroid sensitive (Scotinomys teguina). Males sing more often than females, and on average male songs have more notes. Song is influenced by circulating androgens, but how such hormonal differences influence the diverse brain regions involved in vocal display is not understood. To characterize androgen-sensitive sites in the vocal motor pathway, we used two isoforms of pseudorabies virus (PRV) to double-label circuits ending in laryngeal and jaw muscles involved in vocalization, and co-labeled these neurons for androgen receptor (AR). Next we manipulated circulating androgens and observed the effects on AR distribution and male song. We find androgens drive coordinated changes in AR abundance across motor and motivational circuits, and both individual and group differences in song are associated with AR abundance. The results reveal how circulating androgens and the auto-regulation of androgen receptors can influence the diverse circuits necessary for elaborate advertisement displays. Significance statementCourtship and aggressive displays are among those most elaborate and dramatic of sexually dimorphic behaviors. We show that in Alstons singing mouse, an androgen-sensitive network defined by retrograde tracing shapes vocal display effort. Our results reveal how the intensity of singing mouse vocalizations is influenced by androgen actions in the vocal motor pathway.

neuroscience↗