Biology subjects

Martin, L.

Publications and source records attributed to Martin, L..

At least 19 recordsLinked to original sources

bioRxiv2026

Multimodal cell communication networks nominate immunotherapies for RCC subgroups with discrete T cell recruitment or expansion

Renal cell carcinoma (RCC) is amongst the most immune-infiltrated solid tumours, but only a small subset of patients achieves durable response to immune checkpoint blockade therapy. Efforts to characterize the immune microenvironment and molecular regulators responsible for treatment responses have explored numerous facets of disease biology using compartmentalized genomics, transcriptomics, and proteomics datasets, yielding many important yet context and data specific insights. Therefore, to provide a more integrated approach to informing future precision medicine strategies, we combined the complementary strengths of multiple technological platforms to profile multi-regional, spatially annotated surgical biospecimens from 65 RCC patients by single-cell RNA sequencing with paired TCR and BCR repertoire analysis, imaging mass cytometry, suspension mass cytometry, spatial transcriptomics and deconvolved bulk RNA sequencing. With this resource dataset, we explored patient subgroups and precision immunotherapy strategies using an integrated analysis of transcripts and proteins across single cell and spatial modalities. Proximal cell interactions and distinct receptor-ligand pairings identified 7 recurrent cellular communication networks. Robustly mapping reproducible gene signatures across technologies and to a variety of publicly available datasets, we show these highly refined immune subgroups stratify patients with tumour microenvironments associated with prognosis and immunotherapy response. Notably, this reveals that highly infiltrated environments with the potential for immunotherapy response may in fact comprise two distinct communication networks, with differing modes of T cell clonal expansion and immune evasion axes associated with T cell exhaustion or myeloid and NK reprogramming, which could inform targeted combination therapeutic strategies to improve outcomes. Overall, we provide a high-dimensional multi-modal resource dataset that enables cross-platform integration, links stages of T cell clonal expansion with enabling or suppressive RCC immune cell communication networks and nominates rational strategies for combinatorial precision immunotherapy. (Funded by University Health Network, Toronto; REMEDY ClinicalTrials.gov number, NCT04005183.)

cancer biology↗

bioRxiv2026

Humanized patient-derived xenografts preserve tumour-specific immune microenvironments

Defining the genetic and cellular programs that allow solid tumours to evade immune control requires preclinical models that preserve the complexity of the human tumour immune microenvironment. Most available systems capture only part of this biology. Organoid cultures and ex vivo tumour fragments can retain patient-derived tumour architecture and associated immune cells, but immune populations are typically maintained only for short periods. These models also cannot capture antitumour immune responses in the physiological setting of a living organism. Patient-derived xenografts propagated in humanized mice offer a potential path to overcome these limitations by combining patient-derived tumour tissue with a reconstituted human immune system. However, few studies have systematically tested whether these models reproduce the diverse immune cell phenotypes present in the parental tumours from which they are derived. This has limited their use for studying tumour-intrinsic mechanisms that shape immune composition and promote immune evasion. To address this gap, we profiled tumour-infiltrating, splenic, and bone marrow immune cells from ovarian, head and neck, and renal PDX models propagated in CD34+ hematopoietic stem cell (HSC)-derived huNOG-EXL mice expressing human IL-3 and GM-CSF. By comparing tumours grown across distinct HSC donor backgrounds with their matched primary tumour samples, we found that tumour-intrinsic factors are a dominant determinant of immune composition in humanized PDX tumours. Across models, these immune infiltrates generally resembled those of the corresponding parental tumours. These findings support humanized PDX models as a platform for functionally interrogating tumour-intrinsic drivers of immune composition and immune evasion in solid tumours.

cancer biology↗

bioRxiv2026

Dietary emulsifiers and host inflammation synergistically drive genomic evolution of Crohn disease-associated E. coli toward enhanced pathogenicity

Background and AimsThe rising incidence of Crohns disease (CD) in Westernized countries has been linked to changes in diet and increased consumption of food additives, yet the mechanisms by which these factors fuel intestinal inflammation remain unclear. Adherent-invasive Escherichia coli (AIEC), a pathobiont involved in CD pathogenesis, lacks a clear genetic hallmark but exhibits intestinal colonization and virulence traits, raising questions about the evolutionary forces promoting its emergence among select individuals. Here, we investigated how chronic exposure to two common dietary emulsifiers, carboxymethylcellulose (CMC) and polysorbate 80 (P80), along with host inflammation, drives AIEC genomic evolution and pathogenic potential. MethodsWild-type and Il10-deficient mice were monocolonized with AIEC and chronically exposed to CMC, P80, or water. Bacterial isolates were collected and analyzed for genomic diversification, mutations, and phenotype both in vitro and in vivo. ResultsEmulsifiers accelerated AIEC genomic diversification and selected for mutations linked to increased motility, invasion, and pro-inflammatory activity. Moreover, dietary emulsifier-evolved strains displayed a marked fitness advantage in vivo, outcompeting their counterparts in murine hosts, with the greatest advantage observed when evolution occurred under inflammatory conditions. Notably, evolutionary pathways and phenotypic outcomes were shaped by both emulsifier and the hosts inflammatory status, highlighting synergy between diet and host genetics in fostering pro-inflammatory pathobionts. ConclusionThese findings provide an evolutionary framework connecting modern dietary habits to the emergence of pathogenic AIEC strains, and underscore the importance of dietary interventions in individuals at risk for inflammatory bowel disease.

microbiology↗

bioRxiv2026

Environmental Control of Antimicrobial-to-Amyloidogenic Switching in Uperin 3.5 by Surfactant Assembly Dynamics

Antimicrobial peptides (AMPs) that also form functional amyloids exhibit remarkable environmental sensitivity, yet the physicochemical rules governing their structural switching remain unresolved. Here, we investigate how surfactant charge and assembly dynamics regulate the antimicrobial-amyloidogenic transition of Uperin 3.5, a 17-residue amphibian AMP with pronounced conformational plasticity. Using an integrated approach combining all-atom molecular dynamics simulations with circular dichroism and thioflavin T fluorescence assays, we systematically probe the effects of surfactant identity, concentration relative to the critical micelle concentration (CMC), peptide stoichiometry and ionic strength. We show that -helical stabilisation and antimicrobial-like behaviour scale directly with surfactant charge: anionic Sodium dodecyl sulphate (SDS) induces the highest helicity in monomeric Uperin 3.5 ({approx}80-90%), followed by zwitterionic dodecyl-phosphocholine (DPC) ({approx}35-45%), while cationic Cetyltrimethylammonium bromide (CTAB) fails to stabilise secondary structure. This charge-ordered trend is mirrored in oligomer remodelling, with SDS driving the most rapid dissociation of {beta}-sheet tetramers, DPC inducing slower partial disassembly and CTAB exhibiting minimal effect. Above the CMC, micellar environments stabilise amphipathic -helical states and efficiently dissolve amyloid assemblies. In striking contrast, under below-CMC conditions, limited SDS availability combined with peptide crowding promotes cooperative aggregation, where surfactant monomers act as dynamic scaffolds that nucleate N-terminal {beta}-sheet interactions--an effect strongly accelerated by physiological salt. Large-scale simulations reveal mixed /{beta} aggregates whose formation is governed by electrostatic screening and surfactant-mediated co-assembly. Together, these findings establish surfactant charge and assembly state as quantitative, environment-dependent regulators of functional amyloidogenesis in antimicrobial peptides. More broadly, they suggest that controlled modulation of membrane-mimetic environments can be exploited to bias peptides toward antimicrobial or amyloidogenic states, offering conceptual avenues for therapeutic strategies targeting peptide misfolding and neurodegenerative disorders.

biochemistry↗

bioRxiv2026

Generative Deep Learning and Molecular Dynamics Reveal Design Principles for Amyloid-Like Antimicrobial Peptides

Antimicrobial peptides (AMPs) are emerging as promising alternatives to conventional antibiotics, and growing evidence indicates a fundamental link between antimicrobial activity and amyloid-like self-assembly. Many AMPs are known to form amyloid-like fibrils, while several amyloidogenic peptides exhibit intrinsic antimicrobial properties, suggesting shared underlying physicochemical determinants such as amphipathicity, {beta}-sheet propensity, and charge distribution. However, the rational design of peptides that simultaneously encode these dual functionalities remains a significant challenge. Here, we present amyAMP, a generative deep-learning framework based on a Wasserstein generative adversarial network with gradient penalty (WGAN-GP), designed to learn and generate peptides with integrated antimicrobial and amyloidogenic properties. Trained on curated datasets of antimicrobial and amyloid-forming peptides, amyAMP captures the latent sequence-property relationships governing dual functionality. Statistical and latent-space analyses demonstrate that the generated peptides closely overlap with biologically relevant peptide space while remaining distinct from random sequences, indicating successful learning of key biochemical features. To validate functional behavior, we performed extensive coarse-grained molecular dynamics simulations to probe membrane interaction, peptide self-assembly, and membrane disruption. The simulations reveal rapid membrane adsorption, stable amphipathic insertion, and strong peptide-peptide aggregation. Notably, cooperative clustering of peptides on membrane surfaces induces membrane thinning and curvature perturbations, highlighting a mechanistic coupling between aggregation and antimicrobial activity. Collectively, these results establish that amyAMP effectively captures the shared physicochemical principles underlying antimicrobial action and amyloid-like self-assembly. This work provides a generalizable framework for the AI-guided design of multifunctional peptides to advance the development of next-generation therapeutics targeting antimicrobial resistance.

biophysics↗

bioRxiv2026

Intranasal oxytocin mRNA-LNP can promote social behaviour and reduce pain

The COVID-19 epidemic and success of mRNA-LNP vaccines demonstrated the transformative potential of mRNA therapeutics. Beyond vaccination, mRNA delivery offers a platform for transient on-demand expression of therapeutic proteins for both rare and common diseases. While delivery of therapeutics to the liver is relatively straightforward, targeted delivery of mRNA-LNPs to the central nervous system (CNS) remains a significant challenge. Here we show that intranasal mRNA-LNP delivery results in localized mRNA cargo uptake and functional expression in the respiratory and olfactory epithelium, where the encoded cargo protein is secreted and can enter the CNS. Guided by genomic data of pain-associated gene expression, we identified secreted proteins as candidate mRNA-encoded analgesics. Intranasal mRNA-LNP encoding a synthetic oxytocin transcript (OXT) resulted in bioactive oxytocin peptide delivery to the CNS. Functionally, intranasal OXT mRNA-LNP enhanced social behaviour and attenuated pain responses across multiple behavioural paradigms, without impairing motor coordination. Importantly, repeated dosing was well tolerated and intranasal mRNA-LNP did not elicit an inflammatory response or alter overall health. Together, these findings establish intranasal mRNA-LNP delivery of secreted ligands as a safe, non-invasive route to target the CNS, unlocking a new class of mRNA therapeutics for pain or other disorders of the brain. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=189 SRC="FIGDIR/small/711938v1_ufig1.gif" ALT="Figure 1"> View larger version (39K): org.highwire.dtl.DTLVardef@131252aorg.highwire.dtl.DTLVardef@17efb79org.highwire.dtl.DTLVardef@1afe479org.highwire.dtl.DTLVardef@c0f2f5_HPS_FORMAT_FIGEXP M_FIG C_FIG One sentence summaryIntranasal administration of mRNA-LNP enables local transfection in nasal epithelium and subsequent secretion of therapeutic oxytocin peptides into the brain, promoting social interactions and reducing pain.

neuroscience↗

bioRxiv2026

Human decellularized extracellular matrix from adipose tissue is a permissive microenvironment for pancreatic organoids generation

In vitro reconstruction of human tissue microenvironments that integrate native biochemical and biomechanical cues is essential for disease modelling, regenerative medicine, and personalized therapeutic approaches. However, most currently available engineered matrices fail to recapitulate the complexity and tissue specificity of the human extracellular matrix (ECM). To address this limitation, we developed a novel hydrogel derived from decellularized human adipose tissue (atdECM) designed to support three-dimensional culture of human cells. The decellularization and delipidation processes were first validated, and the biochemical composition and biomechanical properties of atdECM were comprehensively characterized. Human pancreatic organoids were then cultured within atdECM hydrogel, and their structural organization and transcriptional profiles were analyzed and compared with those obtained in Matrigel, the current gold-standard matrix for organoid culture. Proteomic and cytokine analyses demonstrated efficient decellularization while preserving collagen-rich ECM architecture and a diverse repertoire of soluble bioactive factors. AtdECM exhibited physiological stiffness and retained tissue-specific extracellular cues. Pancreatic organoids cultured in atdECM displayed morphological similarities with those grown in Matrigel but exhibited transcriptional profiles more consistent with physiological epithelial homeostasis, with reduced activation of inflammatory and stress-related pathways. Altogether, these findings indicate that atdECM provides a human-derived, tissue-relevant, and permissive microenvironment for human organoid generation. This platform represents a promising alternative to Matrigel for studying human tissue biology and for developing physiologically relevant in vitro models.

bioengineering↗

bioRxiv2026

Physiological cerebrospinal fluid like medium reveals autophagy dependency of leukaemia in the central nervous system

Nutrient availability is a critical environmental factor that influences the metabolism and adaptability of cancer cells, including acute lymphoblastic leukaemia (ALL) cells, prone to relapse in the central nervous system (CNS). Currently available cell culture media contain supraphysiological nutrient levels and do not represent the restricted metabolic environment of CNS-ALL which resides in the leptomeninges surrounded by cerebrospinal fluid (CSF). Therefore, we formulated a novel physiological CSF-like cell culture medium (CSFmax) that recapitulates the unique metabolite composition of the CSF. Through in vitro and in vivo metabolic and functional studies, we demonstrate that ALL cells cultured in CSFmax rewire their metabolism, closely mimicking the metabolic phenotype of CNS-ALL, including their metabolic activity and redox state. Utilising CSFmax, in comparison to conventional nutrient-rich culture media, we identified an essential role for autophagy in ALL adaptation to the CNS niche. This was evident by increased autophagic activity and selective sensitisation of ALL cells to pharmacological inhibition of autophagy and genetic knockout of Unc-51 Like Autophagy Activating Kinase 1 (ULK1) or autophagy related 7 (ATG7). Importantly, using a robust preclinical in vivo model, mice xenografted with ULK1 and ATG7 deficient ALL cells exhibited reduced CNS disease burden when compared to mice xenografted with control cells. Overall, our findings provide strong evidence that physiological CSFmax is superior to current in vitro culture systems in recapitulating the metabolic signature of CNS resident ALL cells. By exploiting this system, we revealed for the first time autophagy as a targetable therapeutic vulnerability in CNS-ALL. Key PointsO_LICulturing ALL cells in bespoke CSF-like medium (CSFmax) recapitulates the metabolic adaptation of ALL cells in the CNS niche C_LIO_LIAutophagy is critical for metabolic adaptation and survival of CNS resident ALL cells C_LI

cancer biology↗

bioRxiv2026

A Stable Reporter Cell Line to Study Respiratory Syncytial Virus NS2-Mediated Inhibition of IFNB Promoter Activation

Human respiratory syncytial virus (RSV) is an enveloped, non-segmented, negative-sense RNA virus and a major cause of severe lower respiratory tract disease in infants, older adults, and immunocompromised individuals. A hallmark of RSV pathogenesis is early evasion of innate immunity mediated by the non-structural proteins NS1 and NS2. NS2 is a multifunctional interferon (IFN) antagonist that dampens both IFN{beta} induction and type I IFN responsiveness by targeting multiple nodes of the RIG-I/IFN signaling axis. Because NS2 inhibition is typically partial and context dependent, its quantitative assessment in conventional cell-based assays can be challenging. Here, we report the generation and characterization of a stable, virus-free reporter system in human lung epithelial A549 cells to measure NS2-mediated inhibition of RIG-I-dependent activation of the human IFNB promoter. The platform combines stable NS2 expression with genomic integration of a dual cassette encoding (i) a non-targeting short hairpin RNA that constitutively activates RIG-I signaling and (ii) a firefly luciferase reporter driven by the IFNB promoter. This configuration provides a sensitive and reproducible luminescence readout of IFNB promoter activity under controlled, infection-free conditions, enabling robust quantification of NS2 antagonism with reduced experimental variability. The luminescence-based format is readily scalable for early-stage discovery and prioritization of small molecules or biologics that counteract NS2 function. More broadly, the modular design can be adapted to other viral immune antagonists, signaling pathways, or cell types, supporting comparative studies where pathway modulation must be evaluated under highly standardized and reproducible conditions.

microbiology↗

bioRxiv2026

A Fine-Tuned Phosphatidylinositol Profile Contributes to Colonocyte Differentiation and Malignization: Evidence From Integrated Omics

Membrane lipid composition changes concomitantly with human colonocyte differentiation, a tightly regulated process occurring along the colon crypt. This process is heavily disrupted in colon cancer. Nonetheless, the regulatory mechanisms driving these changes, especially the replacement of arachidonic acid phosphatidylinositol species with monounsaturated fatty acid species, and how they are altered in cancer, remain unknown. To establish the transcriptional networks underlying this remodeling, we integrated transcriptomic and lipidomic profiles of isolated healthy and tumor human colonocytes using system biology approaches; identifying key gene regulatory networks involved in arachidonic acid and eicosanoid metabolism and phosphatidylinositol cycle as significant regulators during differentiation. Consistently, a distinct impact was found on organoid differentiation depending on colonocyte subtype and specific prostaglandin. Remarkably, the shift and associated transcriptomic programs were lost in tumor that heightened phosphoinositide metabolism. Altogether, these results underscore the importance of lipid remodeling in colonocyte stemness maintenance and proper onset of differentiation programs.

systems biology↗

bioRxiv2026

A proline-rich-domain-binding single domain antibody selectively inhibits RNA-induced liquid-liquid phase separation of tau

Liquid-liquid phase separation (LLPS) mediates the formation of biomolecular condensates, which organize cellular processes such as synaptic plasticity and stress response. The neuronal microtubule-associated protein tau undergoes LLPS under specific conditions, regulating synaptic vesicle clustering and microtubule dynamics. In vitro, tau LLPS is induced by cofactors such as polyethylene glycol (PEG) or RNA, mainly via weak multivalent electrostatic interactions. However, the molecular mechanisms governing the formation of tau LLPS, including domain specific contribution, remain unclear. In this study, we used eight single-domain antibodies (VHHs), targeting six distinct short sequences of tau, to explore the mechanisms of tau LLPS in vitro. By combining several biophysical methods, we evaluated the effect of each anti-tau VHH on tau LLPS with two main LLPS inducers, PEG (molecular crowding) and RNA (complex coacervation). With PEG as an inducer, all VHHs targeting tau enhanced tau LLPS formation, regardless of their affinity for tau. With RNA as an inducer, the effect of the VHHs was mixed: VHHs targeting the C-terminal domain promoted condensation, while VHH B1-1, which binds the proline-rich domain (PRD; including residues (221REPKKVAVVRTP232), abolished droplet formation. NMR and surface plasmon resonance confirmed 1 to 1 binding of VHH B1-1 to the PRD, and competition assays with a PRD peptide restored LLPS, demonstrating mechanistic specificity. This result underscores the importance of this region in tau LLPS formation. Our findings provide domain-resolved insights into the regulation of tau LLPS and demonstrate the potential of VHHs as tools to selectively modulate biomolecular condensates in physiological and pathological contexts.

biochemistry↗

bioRxiv2025

Dendrite-targeting OLM interneurons regulate the formation of learning-related CA1 place cell representations

Spatial learning depends on hippocampal CA1 place cell representations, which form rapidly through behavioral timescale synaptic plasticity (BTSP). BTSP is driven by dendritic plateau potentials proposed to arise from the interaction of an excitatory target signal from entorhinal cortex layer 3 (EC3) and inhibitory feedback reflecting the current CA1 population state. However, the cellular source of this feedback has remained unknown. Using two-photon Ca2+ imaging in mice during spatial learning, we found that dendrite-targeting oriens-lacunosum moleculare (OLM) interneurons increased their activity at behaviorally salient locations in a manner consistent with previously described environment-specific CA1 representations and EC3 target signals. Causal manipulations revealed that silencing a genetically defined subset of OLM interneurons late in learning enhanced BTSP and place field formation, whereas activating them early suppressed place field formation. These findings identify OLM interneurons as a key inhibitory feedback element regulating BTSP and the formation of hippocampal representations during learning.

neuroscience↗

bioRxiv2025

Tuning the open-close equilibrium of Cereblon with small molecules influences protein degradation

Most PROTACs and molecular glue degraders currently approved or in clinical trials recruit Cereblon (CRBN) as the ubiquitin E3 ligase. Upon binding ligands and molecular glues, CRBN undergoes a significant structural rearrangement from an open to closed state, defined by the positioning of the thalidomide-binding domain (TBD) with respect to the Lon domain. However, the exact molecular basis for this ligand-induced conformational change and its implication to neo-substrate degradation remain elusive. During our campaign to discover novel CRBN binders, we found hits exhibiting distinct biophysical behaviour from classical thalidomide-based ligands. By combining orthogonal biophysical methods of differential scanning fluorimetry, isothermal titration calorimetry, and small-angle X-ray scattering, supported by X-ray crystallography and cryo-EM structures of ligand-bound complexes, we classify CRBN binders between those that can induce CRBN closure, and those that cannot. Mutational studies identify key residues in the CRBN ligand binding pocket and N-terminal belt that are essential for the Lon and TBD domains to come together in the closed state. Finally, we show that the probability to yield active degrader molecules is greatly influenced by whether binders can or cannot induce CRBN closure. Together, our study reveals new molecular insights into the structural basis for how CRBN open-closed equilibrium is directly modulated by compound binding and impact target degradability by CRBN, with important implications to the design of PROTACs and molecular glue degraders.

biochemistry↗

bioRxiv2025

An Interactive Brain Atlas of Knowledge

Biomedical knowledge about the brain increases every day, with a rapidly growing number of scientific publications, datasets, and software tools. While this informational plethora is not merely comprehensible by human beings, recent developments in information science and computational linguistics aim to make this knowledge programmatically accessible by literature mining. However, integrating these semantic methods into neuroimaging standards remains insufficient, hindering researchers from unraveling their full potential. Therefore, we developed the semantic meta-analysis platform The Virtual Brain adapter of semantics (TVBase) that projects biomedical knowledge preserved in over 36 million scientific articles onto a 3D standardized brain. The literature-mining platform SCAIView was used to extract ontologically defined biomedical entities and their associations with brain anatomy from the PubMed database. By querying a specific concept, the association strength with each anatomical term was calculated using entropy. To project the data onto a standardized brain, we created a unique transformation matrix that links over 800 anatomical terms to voxel coordinates of a parcellated standard brain. This novel method of knowledge projection extracts region-specific information about biomedical concepts from the literature to support translational multi-scale approaches to computational neuroscience. The multi-purpose software framework TVBase is openly available as a Python library. It aims for hypothesis-free neuroimaging pattern interpretation, hypothesis generation, and applications in personalized medicine.

neuroscience↗

bioRxiv2025

The Virtual Brain Ontology: A Digital Knowledge Framework for Reproducible Brain Network Modeling

Computational models of brain network dynamics offer mechanistic insights into brain function and disease, and are utilized for hypothesis generation, data interpretation, and the creation of personalized digital brain twins. However, results remain difficult to reproduce and compare because equations, parameters, networks, and numerical settings are reported inconsistently across the literature, and shared code is often not fully documented, standardized, or executable. We introduce The Virtual Brain Ontology (TVB-O), a semantic knowledge base, minimal metadata standard, and Python toolbox that simplifies the description, execution, and sharing of network simulations. TVB-O offers 1) a common vocabulary and ontology for core concepts and axioms representing current domain knowledge for simulating brain network dynamics, 2) a minimal, human- and machine-readable metadata specification for the information needed to reproduce an experiment, 3) a curated database of published models, brain networks, and study configurations, and 4) software that generates executable code for various simulation platforms and programming languages, including The Virtual Brain, Jax, or Julia. FAIR metadata and provenance-aware reports can be exported from TVB-Os model specification. It hereby enables a flexible framework for adopting new models and enhances reproducibility, comparability, and portability across simulators, while making assumptions explicit and linking models to biomedical knowledge and observation pathways. By reducing technical barriers and standardizing workflows, TVB-O broadens access to computational neuroscience and establishes a foundation for transparent, shareable "digital brain twins" that integrate with clinical pipelines and large-scale data resources.

neuroscience↗

bioRxiv2025

Fast and Easy Whole-Brain Network Model Parameter Estimation with Automatic Differentiation

Personalized brain modeling at clinically relevant scales requires integrating biophysical models with empirical neuroimaging data, yet high-dimensional parameter estimation in whole-brain network models remains computationally prohibitive. We present TVB-Optim, an open-source Python library providing a general and extensible framework for gradient-based optimization of brain network models build on JAX. Leveraging automatic differentiation, we demonstrate direct optimization of thousands of parameters, from global coupling (N=2) to regional dynamics (N=168) to full structural connectivity matrices (N=14,028), across functional MRI and magnetoencephalography data. Forward simulations achieve 10x CPU speedup over reference implementations and scale efficiently to GPU and multi-device configurations. We establish best practices through documented workflows that combine coarse parameter space exploration with gradient refinement, often yielding superior solutions faster than gradient methods alone. By bridging mechanistic neuroscience with modern machine learning infrastructure, TVB-Optim enables large-scale personalized brain models, bringing computational neuroscience closer to clinical translation.

neuroscience↗

bioRxiv2025

Optimized quantitative bacterial two-hybrid (qB2H) for protein-protein interaction assessment

Characterizing mutation effects on protein-protein interactions (PPIs) is crucial for elucidating protein structure and function. Massively parallel PPI variant analyses such as deep mutational scanning (DMS) enable interface identification and generate datasets for machine learning. In cellulo strategies such as two-hybrid systems provide straightforward access to such data, but reliability depends on quantitative properties. Here, we show that existing bacterial two-hybrid (B2H) systems have limitations constraining accurate dataset generation. We engineered and benchmarked optimized quantitative B2H (qB2H) alternatives, enabling strain-independent assays, improved metrics, and generation of high-quality datasets. We demonstrate qB2H utility through interface mapping and binder optimization. Perturbation analysis of single-site variants accurately recovered known ASF1 complex contact positions, matching crystallographic data. Integration of generative AI-based design yielded an ASF1-binding peptide with a 70-fold increase in affinity. qB2H offers to R&D scientists a robust, reusable platform for quantitative PPI analysis, enabling both rational protein engineering and data-driven discovery.

synthetic biology↗

bioRxiv2025

Strain-resolved CRISPRi-Seq reveals conserved antibiotic vulnerabilities in Staphylococcus aureus

Staphylococcus aureus remains a major clinical threat due to rising antibiotic resistance and high rates of treatment failure. Deciphering the genetic responses to antibiotic pressure and identifying conserved vulnerabilities are essential steps toward developing broadly effective therapies. Here, we constructed strain-resolved CRISPR interference (CRISPRi) libraries targeting all genes in four clinically relevant S. aureus strains spanning major clonal complexes. CRISPRi-seq screens enabled high-resolution mapping of their fitness landscapes and the definition of a core essentialome representing robust targets for antimicrobial intervention. Exposure of the CRISPRi libraries to four mechanistically distinct antibiotics revealed genome-wide susceptibility profiles, identifying both strain-dependent and conserved susceptibility signatures shaped by the drug mode of action and genetic background. Analysis of these conserved vulnerabilities provided insight into antibiotic-specific stress responses and resistance mechanisms. Among the core determinants of vancomycin vulnerability, we identified several previously uncharacterized genes, including a conserved membrane-associated operon, here designated EsrABC, whose disruption markedly increases vancomycin sensitivity in the four strains. Our study provides a genome-wide atlas of S. aureus fitness and conditional vulnerabilities, fully explorable in the here-developed online AureoBrowse platform (https://aureobrowse.veeninglab.com/), revealing candidates for synergistic therapies and potential therapeutic targets.

microbiology↗