Search NASA⌕ Search

SEARCH · Search NASA

Results for “model fine-tuning”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2026

Residual symmetries and scalar multiplet vacuum alignment in non-Abelian flavour models

We demonstrate that, upon minimizing a renormalizable, single-scalar potential invariant under a non-Abelian symmetry, special orientations in the associated vacuum alignment of the scalar multiplet correspond to the preservation of a discrete residual flavour symmetry in the broken phase of the theory. Conversely, we show that these special scalar alignments are perturbed when additional Lagrangian operators (e.g. renormalizable, multi-flavon operators and/or effective, higher-dimensional operators) are present that break said residual symmetry, leading to a vacuum reorientation and phenomenological consequences. We therefore construct a one-to-one correspondence principle between broken residual symmetries and vacuum alignment corrections, providing a mechanism to identify (and correct) a subtle but persistent form of phenomenologically relevant fine-tuning embedded in — but often ignored by — most successful non-Abelian flavour models. We first establish this correspondence in a set of toy models based on the S4 permutation symmetry, and then apply the lessons learned to the more realistic A4 Altarelli-Feruglio and ∆(27) Universal Texture Zero models.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2026

The price of a large electron Yukawa modification

The theoretical implications of an electron Yukawa modification are considered in the context of a possible Higgs pole run at FCC-ee, aimed at bounding this coupling. We start from an effective field theory viewpoint, considering the impact of renormalisation group effects on related observables and also examining assumptions on the broader UV flavour structure. We then give an overview of the landscape of simplified models, investigating phenomenological constraints arising at higher orders. A short discussion of fine-tuning is also included.

Anomalous Higgs Couplings↗
DOE OSTI2026

HydraGNN_GFM_FineTuning4Materials v1.0

This repository enables fine-tuning of the HydraGNN Predictive GFM 2026 — an open-source ensemble of pre-trained graph foundation models for atomistic materials modeling, developed at Oak Ridge National Laboratory. The GFM 2026 is freely available and downloadable via Globus from the OLCF Data Constellation (DOI: 10.13139/OLCF/2562660). Starting from these pre-trained weights, this repository provides a complete transfer learning pipeline for adapting the GFM ensemble to domain-specific molecular and materials property prediction tasks. It includes: 1) Utilities for ensemble fine-tuning with task-specific output heads 2) Example pipelines for eight widely-used materials and molecular datasets 3) Tools for model adaptation and head configuration 4) Data preprocessing utilities for each supported dataset 5) Benchmarking and evaluation scripts

Ungerboeck, Linda↗
DOE OSTI2026

Describing Point Defect Topology in 2D Energy Materials Through Computer Vision

Point defects such as vacancies and impurity atoms strongly impact the performance of 2D materials. Traditional efforts often rely on manual detection, a process that is time-intensive, prone to human error, and challenging to scale. Here we leverage machine learning (ML) methods to identify and quantify vacancies within 2D transition metal carbides (Ti3C2, MXenes), aiming to expedite detection while improving accuracy. MXenes exhibit valuable defect-defined electrochemical properties, but we currently lack statistical understanding of defect topology needed to fully harness these materials. Here we employ a convolutional neural network for semantic segmentation of experimental MXene images, opening an opportunity to conduct a rigorous statistical study on defect hierarchy while investigating local relaxation in the lattice. We show how the integration of ML can yield fundamental insight into point defects, providing a powerful tool that will play an increasingly crucial role in the future of materials science. ML is often not just a matter of straightforward application, and pretrained models proved ineffective in this case. Instead, we trained our own neural network (NN) and applied data augmentation techniques and fine-tuning to the training dataset. Since labeled microscopy data is often scarce, we developed training data from a previously published wide-frame MXene image, using customized Gaussian fitting to locate atomic positions. Our trained model was then applied to a large dataset of experimental images, enabling a statistical study of defect configurations across three samples prepared with different HF etchant concentrations (5%, 9.1%, and 12.5%), as shown in Fig. 1. This also allowed us to investigate local strain around vacancies, though we find that we are limited by the precision of measurements using high-angle annular dark field (HAADF) images, as shown in Fig. 2. This study demonstrates how ML enables large-scale, quantitative analysis of atomic defects - an otherwise infeasible task with traditional methods. While our NN was specialized for Ti3C2 MXenes, the pipeline we developed provides a foundation for future ML models tailored to other materials. Ultimately, we envision embedding the NN onto the microscope to give real-time feedback to the user. To make this a reality, continued work is necessary to fully understand the NN's capabilities and limitations. This study gets one step closer to our goals of automated experimentation moving away from traditional methods of manual labeling. As ML capabilities advance, we hope to continue adapting and applying these techniques in microscopy.

2D materials↗
DOE OSTI2026

Data‐Efficient Generation of Synthetic Microstructures of Polymer‐Bonded Energetic Material With Fine‐Tuned Stable Diffusion

Among current deep learning approaches for synthetic image generation, diffusion-based models stand out in terms of algorithmic stability and ability to retain high-fidelity image features with detailed resolution. Here, in this work, we employ Dreambooth, a method for fine-tuning Stable Diffusion, on X-ray CT images of microstructure of the polymer-bonded form (PBX) of a commonly used high explosive, Pentaerythritol tetranitrate (PETN), which yields generative models for creating synthetic PBX images. The models developed here represent five classes (or ‘lots’) of microstructures and demonstrate successful generation of images of each class with high fidelity, as verified by computed classification accuracy of ∼ 94% or higher. Data augmentation afforded by such image synthesis can be used to more reliably decipher underlying statistics, build processing-structure correlations, recognize off-normal structural anomalies, and identify age-related changes. Ideas related to converting image data into appropriate density mapping and performing mesoscale simulation or surrogate modeling of detonation are also discussed.

Dreambooth↗
DOE OSTI2026

Discrete versus continuous: Enhancing battery optimization in capacity expansion models

This study compares two battery modeling approaches for capacity expansion models: discrete-duration and continuous-duration formulations. In the discrete approach, battery duration is fixed, and power capacity is optimized. In the continuous approach, both power and energy capacities are decision variables, allowing storage duration to be optimized endogenously. Although both discrete-duration and continuous-duration battery formulations are used in long-term power system planning models, the literature has provided limited direct, systematic comparisons of their implications within a common modeling framework. To address this gap, this study implements both approaches in the Regional Energy Deployment System (ReEDS TM ) capacity expansion model using two resource adequacy methods, across a range of future system conditions, and with varying battery cost projections. Results show continuous-duration and high-resolution discrete approaches produce similar capacity expansion outcomes. The continuous formulation achieves faster runtimes compared to discrete-duration runs with many discrete-duration options. However, the discrete-duration approach allows users to choose to have limited fidelity for storage duration options, which in some cases can outperform the continuous formulation. The continuous formulation has the lowest overall system costs, indicating its ability to fine-tune storage duration to better meet specific system needs. This study's findings provide a side-by-side evaluation of discrete and continuous battery modeling approaches and offer guidance for improving the representation of real-world systems, flexibility, and computational efficiency for representing energy storage in long-term power system planning models.

25 ENERGY STORAGE↗
DOE OSTI2026

Self-Admitted Technical Debt in Scientific Software: Prioritization, Sentiment, and Propagation Across Artifacts

Self-admitted technical debt (SATD) impairs scientific software (SSW), yet its prioritization, sentiment, persistence, and propagation remains underexplored. Understanding how SSW developers express, and address SATD is crucial for improving SSW maintenance, and tooling. This study investigates how SATD types and artifacts in SSW are prioritized, how sentiment relates to urgency, SATD removal and resolution rates, and the extent to which SATD propagates across artifacts. We analyzed nine SSW repositories using a SATD classification model and a semantic embedding-based prioritization heuristic. SATD was examined across multiple artifacts, with sentiment assessed via a fine-tuned transformer. Propagation was traced, priority scores compared to static analysis, and removal and resolution rates quantified. SATD in comments, commits, and pull requests receive higher priority than SATD in issues, with negative sentiment amplifying urgency. Resolution and removal rates lag behind open-source software (OSS) averages. Most SATD remains confined to the originating artifact, but longer propagation chains are rare and correlate with higher priority, highlighting persistent and high impact debt. Prioritization is influenced by artifact type and sentiment, while low removal and resolution rates signal persistent debt. Cross-artifact propagation marks high priority, unresolved SATD, providing empirical guidance for targeted monitoring, review prioritization, and tool supported maintenance in SSW.

Melin, Eric [Boise State University]↗
DOE OSTI2026

CTRL-STEER: Closed-Loop Neuron Activation Control in Vision-Language-Action Models

Vision-Language-Action (VLA) models enable test-time behavioral steering via neuron-level interventions, but existing methods use fixed strengths and operate in open loop. This static modulation fails under evolving task dynamics, leading to overcorrection, oscillations, and reduced task success—especially for temporal attributes like speed. We propose CTRL-STEER, a control-theoretic framework that casts activation steering as closed-loop feedback with adaptive, time-varying interventions. Instead of assuming neurons encode temporal concepts, we steer along motion-aligned residual directions and regulate intervention magnitude via feedback. We instantiate this with both PID and reinforcement learning controllers that jointly optimize concept adherence and task success. Experiments on fine-tuned OpenVLA policies across four LIBERO suites show improved stability and a better steering–success trade-off over fixed-coefficient baselines, without retraining the base model.

Babu, Abhijith [Florida International University, ↗
DOE OSTI2026

SAM Code Enhancements for Fission Product Tracking of Noble Gases and Metals in MSRs

This report documents fiscal year 2026 enhancements to the System Analysis Module (SAM) for modeling fission product transport in liquid-fueled molten salt reactors (MSRs). The work advances three principal areas: noble gas transport, noble metal deposition, and user interface improvements. The noble gas transport capability integrates drift-flux gas transport, Henry’s law two-film interphase mass transfer with pressure-based nucleation suppression, Knudsen-regime pore diffusion into porous graphite with a conjugate salt-graphite interface constraint, built-in material properties, five Sherwood-number mass transfer correlations including three derived from high-fidelity NekRS simulations, and xenon-135 reactivity feedback through SAM’s point-kinetics model. This work also presents a comprehensive verification test suite, including new analytically verified cases for pressure-dependent onset of interphase gas transfer in a stagnant vertical pipe, a postulated FLiBe-graphite Xe extraction permeator, a gravity riser with a fission-product source, and a descending pipe with gas redissolution driven by hydrostatic pressure. A machine learning framework for bubble rise velocity prediction in molten salt systems is developed and benchmarked on molten-salt and diverse aqueous bubble datasets. The best-performing fine-tuned transfer-learning networks achieve an 82% reduction in RMSE relative to the Clift correlation, and is implemented directly in SAM. The noble metal transport capability is developed, including a liquid-wall deposition model and a gas-surface flotation mechanism that transfers insoluble particles entrained by sparging gas to wetted structures. Verification tests and demonstration cases cover the surface deposition, flotation efflux, and flotation shedding. Finally, a new [SpeciesTransport] input structure replaces positional global vectors with selfcontained, order-independent, named species blocks, simplifies the specification of multiphase species and decay chains, and remains fully compatible with existing SAM input files. Together, these developments improve the physical fidelity, verification basis, and usability of SAM for system-level analyses of fissionproduct behavior in MSRs.

Mui, Travis (ORCID:0000000303736470)↗
DOE OSTI2026

From Rules to Reasoning: A Survey of Large Language Model-Based Approaches to Scientific Hypothesis and Idea Generation

Scientific hypothesis generation represents a fundamental challenge in contemporary research due to exponentially expanding literature volumes and increasing disciplinary specialization. Large language models (LLMs) have emerged as transformative tools for automated scientific discovery, moving beyond traditional rule-based and literature-mining approaches. Four paradigmatic approaches define current LLM-driven hypothesis generation: direct prompting and fine-tuning methods, knowledge-enhanced frameworks integrating retrieval-augmented generation (RAG), multi-agent collaborative systems simulating research teams, and reasoning-focused approaches implementing cognitive architectures. Domain-specific applications demonstrate statistical equivalence to human expert performance in social psychology, experimental validation in biomedical research, and near-expert quality in astronomy. Evaluation methodologies encompass human expert assessment, LLM-as-judge frameworks, and comprehensive benchmarking systems. Technical challenges include hallucination management, knowledge integration limitations, and balancing novelty with feasibility. Future directions emphasize hybrid neural-symbolic architectures and sophisticated human-AI collaboration models for responsible scientific discovery acceleration.

AI-driven discovery↗
DOE OSTI2026

Selective Amnesia using Contrastive Subnet Erasure for Class Level Unlearning in Vision Models

We study concept-level forgetting in pretrained vision models: removing an entire semantic category so the system no longer recognizes that object in unseen images and contexts, rather than merely forgetting specific training examples. Prior work either applies blunt global projections or fine-tunes parameters, which can introduce collateral damage to unrelated features, add compute, and become unstable as forgetting strength increases. We introduce Contrastive Subnet Erasure (CSE), a training-free, encoder-centric edit that targets a compact set of channels most responsible for the class and attenuates them in a calibrated manner. The modification is algebraically folded into the subsequent layer, yielding no inference-time overhead and leaving task heads unchanged. To evaluate whether forgetting generalizes beyond the data used to specify the class, we introduce a cross dataset protocol in which the class is defined on a source dataset and performance is measured on a disjoint target dataset drawn from a different distribution with no shared images. This setup tests whether the model still fails to recognize the object when it looks different or appears in new scenes, and it helps avoid overfitting to patterns in the source dataset. Across CIFAR 10, CIFAR 100, and ImageNet under this protocol, CSE achieves stronger forgetting of the target class while better preserving non target utility than existing baselines in both single class and multi class settings. Overall, CSE provides a simple, stable, and deployment-ready mechanism for class-level unlearning in vision.

Kotevska, Olivera [ORNL] (ORCID:0000000316772243)↗
DOE OSTI2026

Primordial black hole dark matter: A quantitative parameter sensitivity comparison across formation mechanisms and particle candidates

Primordial black holes (PBHs) in the asteroid-mass window ( 10 17 – 10 22 g ) can account for all of the dark matter without violating any observational constraint, yet are routinely dismissed as fine-tuned. I put that dismissal to the test by applying three complementary sensitivity measures uniformly across a broad landscape: three noninflationary PBH production mechanisms, six classes of inflationary PBH models, and seven particle dark matter benchmarks, all evaluated against the same observable target. Three distinct naturalness universality classes emerge, determined entirely by the analytic structure of the abundance map rather than by the nature of the dark matter candidate. Biased-domain-wall PBHs, in their least model-dependent (free- V b ) form, have the same low sensitivity, Δ = 4.5 , as off-resonance weakly interacting massive particles and freeze-in particles ( Δ = 2 ), a sensitivity that, because it is constant over the entire parameter space of the construction, also coincides trivially with its own Wilson-normalized average within that parameter space (Section Definition and conventions), an equivalence that concerns only the space over which Δ is computed and is not a naturalness statement about the construction as a whole; a further reduction to Δ = 2 is possible only under the additional, independently motivated but not required, assumption that the domain-wall bias is generated by Planck-suppressed operators; early matter-domination PBHs occupy an intermediate tier alongside coannihilating weakly interacting massive particles (WIMPs), unified by a structural identity in which the sensitivity measure equals the logarithm of the ratio of the formation scale to the matter–radiation equality scale; first-order phase transition PBHs, once the more accurate super-exponential collapse probability is used in place of the single-exponential approximation, instead belong to the same highly sensitive tier as resonant WIMP annihilation and single-field inflationary collapse, for a structurally distinct reason; single-field ultraslow-roll inflationary collapse is severely tuned for a distinct reason: a double exponential in which the power spectrum amplitude is itself exponentially sensitive to the inflaton potential coefficients, on top of the exponential collapse sensitivity of the abundance map. My main conclusion is that the claim that PBH dark matter is generically fine-tuned conflates the worst case with a landscape spanning every naturalness tier. The Barbieri-Giudice sensitivity computed here and the Wilson-normalized measure of Iovino and Riotto answer distinct and mutually consistent questions about the same construction, a distinction I clarify within the two-layer decomposition.

Profumo, Stefano [University of California, Santa ↗
DOE OSTI2026

Conserved Macromolecular Architecture of Poplar Secondary Cell Walls Revealed by ssNMR and Atomistic Modeling

The macromolecular architecture of plant secondary cell walls governs wood's mechanical and biochemical properties, yet its natural intra-species variability remains poorly characterized. Here, we combined 13C solid-state NMR (ssNMR), multivariate statistical analysis, and molecular modeling to profile nanoscale structure across 13 genetically diverse Populus trichocarpa genotypes grown in 13C-enriched atmospheres. SsNMR-derived phenotypes spanning composition, structure, mobility, and inter-polymer proximities reveal a conserved architecture, with a subtle yet coordinated variation organizing into dominant structural and secondary mobility axes. A representative atomistic model captures these features and reproduces experimental metrics. Molecular dynamics simulations support a weak but consistent positive correlation between cellulose abundance and crystalline-like order, with interior cellulose chains enriched in tg (trans-gauche) conformations without expanding crystalline cores. Together, experiment and simulation reveal a genetically buffered, broadly conserved nanoscale architecture across genotypes, where subtle fine-tuning of cellulose bundling and matrix packing balances mechanical performance with biological function.

09 BIOMASS FUELS↗
DOE OSTI2026

Leveraging large language models to address data scarcity in machine learning for graphene synthesis

Machine learning in experimental materials science faces significant challenges due to the scarcity of data, which are costly and time-consuming to generate, particularly when relying on in-house experiments. Literature data mining offers a potential solution but introduces issues like mixed data quality, inconsistent formats, and non-uniform reporting of synthesis parameters, resulting in partially missing and heterogeneous features across the dataset. Here, we propose data imputation and feature engineering methods that employ pre-trained large language models (LLMs) to enhance machine learning performance on scarce, heterogeneous datasets, demonstrated on graphene CVD synthesis data and the ML-HydPARK hydrogen storage dataset. GPT models perform data imputation via tailored prompting and semantic normalization of inconsistently reported features through embeddings, for example, to harmonize the complex nomenclature of CVD substrates. Beyond yielding more diverse and richer feature representations than traditional methods such as K-nearest neighbors (KNN) and Multivariate Imputation by Chained Equations (MICE), LLM-based data imputation is evaluated against dataset characteristics and prompting strategies. We vary the level of autonomy granted to the LLM, from generic prompting that leverages pre-trained knowledge for autonomous data generation to data-informed prompting that constrains outputs using target-specific information, and demonstrate which level of autonomy yields superior imputation performance across datasets and feature types. The proposed data engineering methods markedly improve downstream performance; for example, in graphene layer number classification using a support vector machine (SVM), binary accuracy increases from 39% to 65% and ternary accuracy from 52% to 72%. Fine-tuning experiments on both datasets show that combining our proposed LLM-based data imputation and feature encoding methods with numerical machine learning predictors outperforms standalone fine-tuned LLM predictors in data-scarce settings. The proposed strategies emphasize data enhancement techniques rather than refining learning architectures or regularizing loss functions, offering a broadly applicable framework for improving machine learning performance on scarce, inhomogeneous datasets.

Chemical vapor deposition↗
DOE OSTI2026

Large language models for transportation research: Methodologies, state of the art, and future opportunities

The rapid rise of large language models (LLMs) is transforming transportation research, with significant advancements emerging between 2023 and 2025, a period marked by the inception and swift growth of adopting and adapting LLMs for various transportation applications. Despite these significant advancements, however, a systematic review and synthesis of the existing literature remains lacking. This paper aims to fill this gap by providing a comprehensive review of the methodologies and applications of LLMs in transportation. We explore key applications, including autonomous driving, travel behavior prediction, and general transportation-related queries, alongside LLM methodologies such as zero- or few-shot learning, prompt engineering, and fine-tuning. From the review, critical research gaps are identified. From the methodological perspective, many of the research limitations can be addressed by integrating LLMs with existing tools and refining LLM architectures. From the application perspective, research opportunities for LLMs to address various transportation challenges are also explored. By synthesizing these findings, this review not only presents the state-of-the-art LLM adoption and adaptation in transportation, but also proposes future research directions as well as insights and recommendations for policymakers and practitioners, paving the way for greater LLM-driven research innovations in transportation in the future.

42 ENGINEERING↗
DOE OSTI2026

Surrogate Neural Architecture Codesign Package (SNAC-Pack)

Neural architecture search (NAS) is a powerful approach for automating model design, but existing methods often optimize for accuracy alone or rely on proxy metrics such as bit operations (BOPs) that correlate poorly with hardware cost. This gap is particularly large for FPGA deployment, where cost is dominated by a multi-dimensional budget of lookup tables, DSPs, flip-flops, BRAM, and latency. We present the Surrogate Neural Architecture Codesign Package (SNAC-Pack), an open-source AutoML framework for hardware-aware neural architecture codesign and end-to-end FPGA deployment. SNAC-Pack runs a multi-objective global search with Optuna and NSGA-II, loading trials to a shared SQLite store that enables parallel workers across compute nodes. A hardware surrogate model outputs per-trial resource and latency estimates, avoiding the synthesis cost that would otherwise dominate the search loop. A local search stage then applies quantization-aware training (QAT) together with iterative magnitude pruning in a combined compression loop, after which the final model is synthesized to FPGA firmware via the hls4ml Python library. A YAML configuration and an optional agentic frontend let users run the pipeline on new datasets without modifying the framework. We demonstrate SNAC-Pack on jet classification at the Large Hadron Collider and superconducting qubit readout, discovering compact architectures that match or exceed strong baselines on the task metric while reducing FPGA resource utilization and, in the qubit readout case, reducing the design space exploration process from months of manual fine-tuning to hours of automated search.

Weitz, Jason [UC, San Diego]↗
DOE OSTI2026

Scalable Generation of High-fidelity Synthetic Population Ensembles

Used within social simulations, synthetic population ensembles enable uncertainty quantification (UQ) methods for obtaining more robust model inference and prediction. A synthetic population ensemble is a series of plausible virtual reconstructions of an area’s population at the granularity of people and residences, generated stochastically to preserve privacy of the source population survey’s respondents. In this paper, we demonstrate the production of large synthetic population ensembles for the U.S. via Oak Ridge National Laboratory’s UrbanPop framework to support modeling of high spatial resolution energy affordability metrics from nationwide social surveys in collaboration with the fusionACS project. The study involves two scenarios: creating ensembles for (1) 17 U.S. metropolitan areas in 2019 and (2) full U.S. Census Divisions in 2023, with each scenario consisting of 41 population instances (a base realization and 40 replicates). To accomplish this task at scale, we configured an integrated system within a research cloud, comprised of virtual containerizations, GPU-enhanced functionality, and orchestrated deployments of UrbanPop’s maturing Likeness Python ecosystem. Results demonstrate we maintained high-fidelity approximations of residential totals by areas of interest and the demographic characteristics of neighborhoods while reducing manual workflow burdens. Finally, we discuss plans to fine-tune and further develop our automated workflows for truly distributed job orchestration to increase computational efficiency, as well as provide an outlook for broadening applications of the ensembles.

Cluster computing↗
DOE OSTI2025

A molecular conveyor belt-associated protein controls the rotational direction of the bacterial type 9 secretion system

ABSTRACT Many bacteria utilize the type 9 secretion system (T9SS) for gliding motility, surface colonization, and pathogenesis. This dual-function motor supports both gliding motility and protein secretion, where rotation of the T9SS plays a central role. Fueled by the energy of the stored proton motive force and transmitted through the torque of membrane-anchored stator units, the rotary T9SS propels an adhesin-coated conveyor belt along the bacterial outer membrane like a molecular snowmobile, thereby enabling gliding motion. However, the mechanisms controlling the rotational direction and gliding motility of T9SS remain elusive. Shedding light on this mechanism, we find that in the gliding bacteriumFlavobacterium johnsoniae, deletion of the C-terminus of the conveyor belt-associated protein GldJ controls and, in fact, reverses the rotational direction of T9SS from counterclockwise (CCW) to clockwise (CW). This suggests that the interface between the conveyor belt-associated protein GldJ and the T9SS ring protein GldK plays an important role in controlling the directionality of T9SS, potentially by modulating its interaction with the stator complex GldLM, which drives motor rotation. Combined with MD simulation of the T9SS stator units GldLM, we suggest a “tri-component gearset” model where GldJ controls the rotational direction of its driver, the T9SS, thus providing adaptive sensory feedback to influence the motility of the gliding bacterium. IMPORTANCE The type 9 secretion system (T9SS) is fundamental to bacterial gliding motility, pathogenesis, and surface colonization. Our findings reveal that the C-terminal region of the conveyor belt-associated protein GldJ functions as a molecular switch which is capable of reversing the rotational direction of T9SS. Through the coordinated actions of the T9SS stator units (akin to a driving motor), the GldK ring (the gear that converts rotational energy into linear movement), and GldJ, this machinery forms a smart conveyor belt system reminiscent of flexible or cognitive mechanical conveyors. Such advanced conveyors can alter their direction to adapt to shifting demands. Here, we show that the bacterial T9SS similarly adjusts its rotational bias based on feedback from the conveyor belt-associated protein GldJ. This dual-role feedback mechanism underscores an evolved, controllable biological snowmobile, offering new avenues for studying how bacteria fine-tune motility in dynamic environments.

Microbiology↗