搜索结果：bioengineering

Effective representations of protein sequences are widely recognized as a cornerstone of machine learning-based protein design. Yet, protein bioengineering poses unique challenges for sequence representation, as experimental datasets typically feature few mutations, which are either sparsely distributed across the entire sequence or densely concentrated within localized regions. This limits the ability of sequence-level representations to extract functionally meaningful signals. In addition, comprehensive comparative studies remain scarce, despite their crucial role in clarifying which representations best encode relevant information and ultimately support superior predictive performance. In this study, we systematically evaluate multiple ProtBERT and ESM2 embedding variants as sequence representations, using the adeno-associated virus capsid as a case study and prototypical example of bioengineering, where functional optimization is targeted through highly localized sequence variation within an otherwise large protein. Our results reveal that, prior to fine-tuning, amino acid-level embeddings outperform sequence-level representations in supervised predictive tasks, whereas the lat

A major challenge for niche scientific and technical domains in leveraging coding agents is the lack of access to up-to-date, domain- specific knowledge. Foundational models often demonstrate limited reasoning capabilities in specialized fields and cannot inherently incorporate knowledge that evolves through ongoing research and experimentation. Materials scientists exploring novel compounds, communication engineers designing and evaluating new protocols, and bioengineering researchers conducting iterative experiments all face this limitation. These experts typically lack the resources to fine-tune large models or continuously embed new findings, creating a barrier to adopting AI-driven coding agents. To address this, we introduce a framework that gives coding agents instanta- neous access to research repositories and technical documentation, enabling real-time, context-aware operation. Our open-source im- plementation allows users to upload documents via doc-search.dev and includes zed-fork, which enforces domain-specific rules and workflows. Together, these tools accelerate the integration of coding agents into specialized scientific and technical workflows

Adeno-associated viral (AAV) vectors are widely used delivery platforms in gene therapy, and the design of improved capsids is key to expanding their therapeutic potential. A central challenge in AAV bioengineering, as in protein design more broadly, is the vast sequence design space relative to the scale of feasible experimental screening. Machine-guided generative approaches provide a powerful means of navigating this landscape and proposing novel protein sequences that satisfy functional constraints. Here, we develop a generative design framework based on protein language models and reinforcement learning to generate highly novel yet functionally plausible AAV capsids. A pretrained model was fine-tuned on experimentally validated capsid sequences to learn patterns associated with viability. Reinforcement learning was then used to guide sequence generation, with a reward function that jointly promoted predicted viability and sequence novelty, thereby enabling exploration beyond regions represented in the training data. Comparative analyses showed that fine-tuning alone produces sequences with high predicted viability but remains biased toward the training distribution, whereas re

Inspired by numerous lab on a chip, biomedical and bioengineering applications such as cell sorting, focusing, trapping, and filtering of particles, manipulation of micron sized particle trajectories has been of significant interest in the context of microfluidics. Systematic deflection of microparticles away from their initial streamlines is a central objective in microfluidic particle manipulation. In many widely used microfluidic platforms including deterministic lateral displacement (DLD) devices, density matched, force free particles suspended in low Reynolds number flows encounter arrays of obstacles that potentially breaks the flow symmetry and alter their trajectories. Despite the prevalence of these devices, the physical mechanism responsible for particle deflection from encountering obstacle wall in strictly non inertial flows (Stokes flows) remains incompletely understood and is often attributed to short range contact interactions rather than hydrodynamic effects.

Modeling relaxation phenomena in complex media is central to understanding multiscale dynamics in materials science, bioengineering and condensed matter physics. Existing fractional-order models, while flexible, sometimes lack physical interpretability, closed-form time-domain expressions, and compatibility with physically realizable architectures. In this work, we propose a novel passive element whose impedance and admittance are defined analytically via modified Bessel functions of first kind, through the electro-mechanical analogy. This approach preserves key physical properties such as analyticity, passivity, BIBO (bounded-input, bounded-output) stability and monotonicity, while enabling the direct use of its time-domain representation in simulations and system modeling. As an application, we demonstrate that this model accurately captures the broadband dispersive behavior of biological tissues, offering a physically grounded and tractable alternative to fractional-order formulations.

Melanoma is the most lethal form of skin cancer, and early detection is critical for improving patient outcomes. Although dermoscopy combined with deep learning has advanced automated skin-lesion analysis, progress is hindered by limited access to large, well-annotated datasets and by severe class imbalance, where melanoma images are substantially underrepresented. To address these challenges, we present the first systematic benchmarking study comparing four GAN architectures-DCGAN, StyleGAN2, and two StyleGAN3 variants (T/R)-for high-resolution melanoma-specific synthesis. We train and optimize all models on two expert-annotated benchmarks (ISIC 2018 and ISIC 2020) under unified preprocessing and hyperparameter exploration, with particular attention to R1 regularization tuning. Image quality is assessed through a multi-faceted protocol combining distribution-level metrics (FID), sample-level representativeness (FMD), qualitative dermoscopic inspection, downstream classification with a frozen EfficientNet-based melanoma detector, and independent evaluation by two board-certified dermatologists. StyleGAN2 achieves the best balance of quantitative performance and perceptual quality,

Biomolecules exhibit a remarkable property of transforming signals from their environment. This paper presents a communication system design using a light-modulated protein channel: Synthetic Photoisomerizable Azobenzene-regulated K+ (SPARK). Our approach involves a comprehensive design incorporating the SPARK-based receiver, encoding methods, modulation techniques, and detection processes. By analyzing the resulting communication system, we determine how different parameters influence its performance. Furthermore, we explore the potential design in terms of bioengineering and demonstrate that the data rate scales up with the number of receptors, indicating the possibility of achieving high-speed communication.

The study of ecological systems is gaining momentum in modern scientific research, driven by an abundance of empirical data and advancements in bioengineering techniques. However, a full understanding of their dynamical and thermodynamical properties, also in light of the ongoing biodiversity crisis, remains a formidable endeavor. From a theoretical standpoint, modeling the interactions within these complex systems -- such as bacteria in microbial communities, plant-pollinator networks in forests, or starling murmurations -- presents a significant challenge. Given the characteristic high dimensionality of the datasets, alternative elegant approaches employ random matrix formalism and techniques from disordered systems. In these lectures, we will explore two cornerstone models in theoretical ecology: the MacArthur/Resource-Consumer model, and the Generalized Lotka-Volterra model, with a special focus on systems composed of a large number of interacting species. In the second part, we will highlight timely directions, particularly to bridge the gap with empirical observations and detect macroecological patterns.

Deep brain stimulation (DBS) is an advanced surgical treatment for the symptoms of Parkinson's disease (PD), involving electrical stimulation of neurons within the basal ganglia region of the brain. DBS is traditionally delivered in an open-loop manner using fixed stimulation parameters, which may lead to suboptimal results. In an effort to overcome these limitations, closed loop DBS, using pathological subthalamic beta (13--30 Hz) activity as a feedback signal, offers the potential to adapt DBS automatically in response to changes in patient symptoms and side effects. However, clinically implemented closed-loop techniques have been limited to date to simple control algorithms, due to the inherent uncertainties in the dynamics involved. Model-free control, which has already seen successful applications in the field of bioengineering, offers a way to avoid this limitation and provides an alternative method to apply modern control approach to selective suppression of pathological oscillations.

Measuring stress fields in fluids and soft materials is crucial in various fields such as mechanical engineering, medicine, and bioengineering. However, conventional methods that calculate stress fields from velocity fields struggle to measure complex fluids where the stress constitutive equation is unknown. To address this, we propose a novel approach that combines photoelastic measurements -- which can non-invasively visualize internal stresses -- with machine learning to measure stress fields. The machine learning model, which we named physics-informed convolutional encoder-decoder (PICED), integrates a convolutional neural network (CNN)-based encoder-decoder model with a physics-informed neural network (PINN). Using this approach, three-dimensional stress fields can be predicted with high accuracy for multiple interpolated data points in a rectangular channel flow.

Gene expression is a complex phenomenon involving numerous interlinked variables, and studying these variables to control expression is essential in bioengineering and biomanufacturing. While cloning techniques for achieving plasmid libraries that cover large design spaces exist, multiplex techniques offering cell culture screening at similar scales are still lacking. We introduced a microcapillary array-based platform aimed at high-throughput, multiplex screening of miniature cell cultures through fluorescent reporters.

Introduction: Taylor & Francis journal Bioengineered has been targeted by paper mills. The goal of this study is to identify problematic articles published in Bioengineered during the period 2010 to 2024. Methods: Dimensions was used to search for articles that contained the terms mouse OR mice OR rat OR rats in title or abstract, published in Bioengineered between January 1st 2010 to December 31st 2024. All articles were assessed by eye and by using software to detect inappropriate image duplication and manipulation. An article was classified as problematic if it contained inappropriate image duplication or manipulation or had been previously retracted. Problematic articles were reported on PubPeer by the authors, if they had not been reported previously. All included articles were assessed for post-publication editorial decisions. Results: We have excluded all articles published in 2024 from further analysis, as these were all retraction notices. We assessed the remaining 878 articles, of which 226 (25.7%) were identified as problematic, of which 35 had been previously retracted. One retracted article was later de-retracted. One article received a correction. None of the incl

Sexual dimorphism is a critical factor in many biological and medical research fields. In biomechanics and bioengineering, understanding sex differences is crucial for studying musculoskeletal conditions such as temporomandibular disorder (TMD). This paper focuses on the association between the craniofacial skeletal morphology and temporomandibular joint (TMJ) related masticatory muscle attachments to discern sex differences. Data were collected from 10 male and 11 female cadaver heads to investigate sex-specific relationships between the skull and muscles. We propose a conditional cross-covariance reduction (CCR) model, designed to examine the dynamic association between two sets of random variables conditioned on a third binary variable (e.g., sex), highlighting the most distinctive sex-related relationships between skull and muscle attachments in the human cadaver data. Under the CCR model, we employ a sparse singular value decomposition algorithm and introduce a sequential permutation for selecting sparsity (SPSS) method to select important variables and to determine the optimal number of selected variables.

Computational protein design (CPD) offers transformative potential for bioengineering, but current deep CPD models, focused on universal domains, struggle with function-specific designs. This work introduces a novel CPD paradigm tailored for functional design tasks, particularly for enzymes-a key protein class often lacking specific application efficiency. To address structural data scarcity, we present CrossDesign, a domain-adaptive framework that leverages pretrained protein language models (PPLMs). By aligning protein structures with sequences, CrossDesign transfers pretrained knowledge to structure models, overcoming the limitations of limited structural data. The framework combines autoregressive (AR) and non-autoregressive (NAR) states in its encoder-decoder architecture, applying it to enzyme datasets and pan-proteins. Experimental results highlight CrossDesign's superior performance and robustness, especially with out-of-domain enzymes. Additionally, the model excels in fitness prediction when tested on large-scale mutation data, showcasing its stability.

Synaptic plasticity dynamically shapes the connectivity of neural systems and is key to learning processes in the brain. To what extent the mechanisms of plasticity can be exploited to drive a neural network and make it perform some kind of computational task remains unclear. This question, relevant in a bioengineering context, can be formulated as a control problem on a high-dimensional system with strongly constrained and non-linear dynamics. We present a self-contained procedure which, through appropriate spatio-temporal stimulations of the neurons, is able to drive rate-based neural networks with arbitrary initial connectivity towards a desired functional state. We illustrate our approach on two different computational tasks: a non-linear association between multiple input stimulations and activity patterns (representing digit images), and the construction of a continuous attractor encoding a collective variable in a neural population. Our work thus provides a proof of principle for emerging paradigms of in vitro computation based on real neurons.

To provide insight into the basic properties of emerging structures when bacteria or other microorganisms conquer surfaces, it is crucial to analyze their growth behavior during the formation of thin films. In this regard, many theoretical studies focus on the behavior of elongating straight objects. They repel each other through volume exclusion and divide into two halves when reaching a certain threshold length. However, in reality, hardly any object of a certain elongation is perfectly straight. Therefore, we here study the consequences of the curvature of individuals on the growth of colonies and thin active films. This individual curvature, so far hardly considered, turns out to qualitatively affect the overall growth behavior of the colony. Particularly, strings of stacked curved cells emerge that show branched structures, while the size of orientationally ordered domains in the colony is significantly decreased. Furthermore, we identify emergent spatio-orientational coupling that is not observed in colonies of straight cells. Our results are important for a fundamental understanding of the interaction and spreading of microorganisms on surfaces, with implications for medical

Designing ligand-binding proteins, such as enzymes and biosensors, is essential in bioengineering and protein biology. One critical step in this process involves designing protein pockets, the protein interface binding with the ligand. Current approaches to pocket generation often suffer from time-intensive physical computations or template-based methods, as well as compromised generation quality due to the overlooking of domain knowledge. To tackle these challenges, we propose PocketFlow, a generative model that incorporates protein-ligand interaction priors based on flow matching. During training, PocketFlow learns to model key types of protein-ligand interactions, such as hydrogen bonds. In the sampling, PocketFlow leverages multi-granularity guidance (overall binding affinity and interaction geometry constraints) to facilitate generating high-affinity and valid pockets. Extensive experiments show that PocketFlow outperforms baselines on multiple benchmarks, e.g., achieving an average improvement of 1.29 in Vina Score and 0.05 in scRMSD. Moreover, modeling interactions make PocketFlow a generalized generative model across multiple ligand modalities, including small molecules, pe