Real‐Time, Label‐Free Classification of Cell Death Pathways via Holotomography‐Based Deep Learning Framework

To evade the immune system, the etiologic agent of plague, Yersinia pestis, injects an exceptionally active tyrosine phosphatase called YopH into host cells using a type III secretion system. We recently reported that YopH acutely inhibits T cell antigen receptor signaling by dephosphorylating the Lck tyrosine kinase. Here, we show that prolonged presence of YopH in primary T cells or Jurkat T leukemia cells causes apoptosis, detected by annexin V binding, mitochondrial breakdown, caspase activation, and internucleosomal fragmentation. YopH also causes cell death when expressed in HeLa cells, and this cell death was inhibited by YopH-specific small molecule inhibitors. Cell death induced by YopH was also prevented by caspase inhibition or co-expression of Bcl-xL. We conclude that YopH not only paralyzes T cells acutely, but also ensures that the cells will not recover to induce a protective immune response but instead undergo mitochondrially regulated programmed cell death.

Objective: To design a deep transfer learning framework for modeling fluence map predictions for stereotactic body radiation therapy (SBRT) of adrenal cancer and similar sites that usually have a small number of cases. Approach: We developed a transfer learning framework for adrenal SBRT planning that leverages knowledge in a pancreas SBRT planning model. Treatment plans from the two sites had different dose prescriptions and beam settings but both prioritized gastrointestinal sparing. A base framework was first trained with 100 pancreas cases. This framework consists of two convolutional neural networks (CNN), which predict individual beam doses (BD-CNN) and fluence maps (FM-CNN) sequentially for 9-beam intensity-modulated radiation therapy (IMRT) plans. Forty-five adrenal plans were split into training/validation/test sets with the ratio of 20/10/15. The base BD-CNN was re-trained with transfer learning using 5/10/15/20 adrenal training cases to produce multiple candidate adrenal BD-CNN models. The base FM-CNN was directly used for adrenal cases. The deep learning (DL) plans were evaluated by several clinically relevant dosimetric endpoints, producing a percentage score relative to the clinical plans. Main results: Transfer learning significantly reduced the number of training cases and training time needed to train such a DL framework. The adrenal transfer learning model trained with 5/10/15/20 cases achieved validation plan scores of 85.4/91.2/90.7/89.4, suggesting that model performance saturated with 10 training cases. Meanwhile, a model using all 20 adrenal training cases without transfer learning only scored 80.5. For the final test set, the 5/10/15/20-case models achieved scores of 73.5/75.3/78.9/83.3. Significance: It is feasible to use deep transfer learning to train an IMRT fluence prediction framework. This technique could adapt to different dose prescriptions and beam configurations. This framework potentially enables DL modeling for clinical sites that have a limited dataset, either due to few cases or due to rapid technology evolution.

Accurate prediction and confirmation of mechanical properties are critical for the reliable design and deployment of sustainable aluminum–agro-waste composites (AWCs). This study presents a confirmatory analytical–experimental–deep machine learning (DML) framework for the identification of true material properties under multi-response conditions. Experimental characterization yielded mean tensile strength of 190.61 MPa, compressive strength of 287.16 MPa, flexural strength of 243.94 MPa, wear rate of 5.46 × 10⁻⁶ mm³ /N·m, coefficient of friction of 0.34, and fatigue life of 5.86 × 10⁵ cycles, with low variability observed in strength-related properties and higher dispersion in tribological and fatigue responses. Isolated analytical predictions produced comparable estimates for tensile (189.60 MPa), compressive (296.42 MPa), flexural (255.75 MPa), and fatigue properties (5.97 × 10⁵ cycles), but failed to reliably capture tribological behavior, yielding a non-physical negative wear rate (− 4.06 mm³/N·m) and an overestimated coefficient of friction (0.45). Consolidated analytical modeling improved agreement for tensile (187.60 MPa), compressive (285.63 MPa), flexural (244.88 MPa), and fatigue life (5.86 × 10⁵ cycles), yet significantly deviated in tribological predictions, producing unrealistically high wear rate (0.20 mm³/N·m) and coefficient of friction (2.84). To resolve these discrepancies, DML models—implemented via a deep convolutional neural network (ID-CNN) and a hybrid residual CNN guided by a residual-based loss function—were employed to reconcile analytical (isolated/consolidated) predictions with experimental data. The DML framework successfully minimized analytical–experimental residuals and confirmed true predictive reliability for compressive strength, flexural strength, coefficient of friction, and fatigue life, while consistently identifying wear rate as falsely predicted across analytical frameworks. These findings establish DML as both a predictive and diagnostic tool, enabling discrimination between truly confirmed and analytically unstable properties. The proposed framework offers a scalable, data-informed pathway for optimized composite design, process refinement, and industrial adoption, with potential extension to multi-scale modeling and real-time process monitoring. Residual-based DML loss quantifies and minimizes analytical–experimental prediction errors. • Residual-based DML loss quantifies and minimizes analytical–experimental prediction errors. • Integrated analytical–DML modeling confirms strength, fatigue, and friction properties. • Bulk properties converge, while wear rate shows persistent tribological uncertainty. • The framework supports reliable certification and sustainable composite design.

The dynactin complex is required for activation of the dynein motor complex, which plays a critical role in various cell functions including mitosis. During metaphase, the dynein-dynactin complex removes spindle checkpoint proteins from kinetochores to facilitate the transition to anaphase. Three components (p150(Glued), dynamitin, and p24) compose a key portion of the dynactin complex, termed the projecting arm. To investigate the roles of the dynactin complex in mitosis, we used RNA interference to down-regulate p24 and p150(Glued) in human cells. In response to p24 down-regulation, we observed cells with delayed metaphase in which chromosomes frequently align abnormally to resemble a "figure eight," resulting in cell death. We attribute the figure eight chromosome alignment to impaired metaphasic centrosomes that lack spindle tension. Like p24, RNA interference of p150(Glued) also induces prometaphase and metaphase delays; however, most of these cells eventually enter anaphase and complete mitosis. Our findings suggest that although both p24 and p150(Glued) components of the dynactin complex contribute to mitotic progression, p24 also appears to play a role in metaphase centrosome integrity, helping to ensure the transition to anaphase.

In this study, a new method for the analysis of harmonic components in the power system based on a deep learning (DL) framework is introduced. In the proposed method, both amplitudes and phases of the harmonic components can be estimated accurately, unlike most of the research work in the literature, which usually focus on estimating amplitudes only. A convolutional neural network (CNN) structure is used to estimate the phases and amplitudes of harmonics, although CNN is usually used for classification. It has been shown that the proposed DL‐based method can satisfactorily estimate both amplitudes and phases of the power system harmonics inside a 20‐ms window and this makes the proposed method suitable for possible real‐time applications, such as active power filtering of the harmonics. It has also been shown that the proposed method is robust to fundamental frequency changes. Experiments on carefully‐generated data sets to reflect the power system behaviour show that the proposed method demonstrates remarkably good performance in terms of estimation accuracy, especially for time‐varying frequency cases. Average error for the amplitude estimation is obtained as 0.21% and that for the phase is 9°, which outperforms the other compared analyses methods in cases of fundamental frequency variations.

The classification of architectural style for Chinese traditional settlements (CTSs) has become a crucial task for developing and preserving settlements. Traditionally, the classification of CTSs primarily relies on manual work, which is inefficient and time consuming. Inspired by the tremendous success of deep learning (DL), some recent studies attempted to apply DL networks such as convolution neural networks (CNNs) to achieve automated classification of the architecture styles. However, these studies suffer overfitting problems of the CNNs, leading to inferior classification performance. Moreover, most of the studies apply the CNNs as a black box providing limited interpretability. To address these limitations, a new DL classification framework is proposed in this study to overcome the overfitting problem by transfer learning and learning-based data augmentation technique (i.e., AutoAugment). Furthermore, we also employ class activation map (CAM) visualization technique to help understand how the CNN classifiers work to abstract patterns from the input. Specifically, due to a lack of architectural style datasets for the CTSs, a new annotated dataset is first established with six representative classes. Second, several representative CNNs are leveraged to benchmark the new dataset. Third, to address the overfitting problem of the CNNs, a new DL framework is proposed which combines transfer learning and AutoAugment to improve the classification performance. Extensive experiments are conducted on the new dataset to demonstrate the effectiveness of our framework. The proposed framework achieves much better performance than baselines, greatly mitigating the overfitting problem. Additionally, the CAM visualization technique is harnessed to explain what and how the CNN classifiers implicitly learn for recognizing a specified architectural style.

Palm oil is a major contributor to the economy, especially in regions like Riau, Indonesia, where it accounts for a significant portion of the Gross Domestic Product (GDP) and national crude palm oil (CPO) production. However, the rising demand for palm oil has led to unstable prices and health concerns, particularly when the oil is reused multiple times. Repeated use of palm oil can increase the risk of health issues such as coronary heart disease and cancer due to the presence of saturated fats and carcinogenic compounds. To address these concerns, the study explores methods to assess palm oil quality, specifically using a fiber optic sensor-based evanescent absorption technique. This method detects changes in the oil by measuring the concentration of substances on the sensor surface without damaging the sample. It operates with high sensitivity, detecting even small changes in the refractive index. Machine learning, particularly Convolutional Neural Networks (CNNs), is used to automate analyzing images and detecting patterns related to palm oil quality. The study being summarized focuses on detecting the quality of palm oil from five different brands of cooking oil in Indonesia. The detection device uses the evanescent absorption method, measuring the voltage of the optical signal converted by a photodiode. The study found that the ratio of initial to measured voltage indicates changes in turbidity and refractive index, which increase as the oil is reused. Among the brands tested, Brand B showed the highest ratio after three uses and a high sensitivity, which is 0.456, indicating it becomes turbid quickly. Brand C has a response time of 67–115 seconds, suggesting it could be reused more times without significant degradation

Bcl-2 is the best characterized member of a large family of proteins that regulate apoptosis. Although it is established that Bcl-2 localized at the mitochondria functions as an anti-apoptotic protein, the function of Bcl-2 at the nucleus remains unclear. Recently we showed that nuclear compartment-associated Bcl-2 inhibits transcription factor activation. Based on this observation, we hypothesized that presence of Bcl-2 at the nucleus may induce rather than protect cells from apoptosis. Here we investigated the putative apoptotic role of nuclear compartment-associated Bcl-2. Additionally, we examined the role of the Bcl-2 BH4 domain in mediating binding to FKBP38, the Bcl-2 mitochondrial chaperone. Our results demonstrate a novel, pro-apoptotic function for nuclear Bcl-2 and identify the Bcl-2 BH4 domain as a key regulator in mediating Bcl-2/FKBP38 binding. These results indicate that Bcl-2 has a dual role as both a protector and a killer and that the ability to switch roles depends on Bcl-2 subcellular localization.

A comprehensive empirical study on bug characteristics of deep learning frameworks

Deep learning (also known as deep structured learning, hierarchical learning, or deep machine learning) is a branch of machine learning based on a set of algorithms that attempt to model high-level abstractions in data by using a deep graph with multiple processing layers, composed of multiple linear and nonlinear transformations. Deep learning has been characterized as a class of machine learning algorithms with the following characteristics [257]: • They use a cascade of many layers of nonlinear processing units for feature extraction and transformation. Each successive layer uses the output from the previous layer as input. The algorithms may be supervised or unsupervised and applications include pattern analysis (unsupervised) and classification (supervised). • They are based on the (unsupervised) learning of multiple levels of features or representations of the data. Higher-level features are derived from lower-level features to form a hierarchical representation. • They are part of the broader machine learning field of learning representations of data. • They learn multiple levels of representations that correspond to different levels of abstraction; the levels form a hierarchy of concepts. These definitions have in common: multiple layers of nonlinear processing units and the supervised or unsupervised learning of feature representations in each layer, with the layers forming a hierarchy from low-level to high-level features. Various deep learning architectures such as deep neural networks, convolutional deep neural networks, deep belief networks (DBN), and recurrent neural networks have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition, and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. In this chapter we start in Section 7.1 with an introduction, giving a brief history of this field, the relevant literature, and its applications. Then we study some basic concepts of deep learning such as convolutional neural networks, recurrent neural networks, backpropagation algorithm, restricted Boltzmann machines, and deep learning networks in Section 7.2. Then we illustrate three examples for Apache Spark implementation for mobile big data (MBD), user moving pattern extraction, and combination with nonparametric Bayesian learning, respectively, in Sections 7.3 through 7.5. Finally, we have summary in Section 7.6.

Cytoplasmic viscosity is a crucial parameter in determining rates of diffusion-limited reactions. Changes in viscosity are associated with several diseases, whilst nuclear viscosity determines gene integrity, regulation and expression. Yet how drugs including DNA-damaging agents affect viscosity is unknown. We demonstrate the use of a platinum complex, Pt[L]Cl, that localizes efficiently mostly in the nucleus as a probe for nuclear viscosity. The phosphorescence lifetime of Pt[L]Cl is sensitive to viscosity and provides an excellent tool to investigate the impact of DNA damage. We show using Fluorescence Lifetime Imaging (FLIM) that the lifetime of both green and red fluorescent proteins (FP) are also sensitive to changes in cellular viscosity and refractive index. However, Pt[L]Cl proved to be a more sensitive viscosity probe, by virtue of microsecond phosphorescence lifetime versus nanosecond fluorescence lifetime of FP, hence greater sensitivity to bimolecular reactions. DNA damage was inflicted by either a two-photon excitation, one-photon excitation microbeam and X-rays. DNA damage of live cells causes significant increase in the lifetime of either Pt[L]Cl (HeLa cells, 12.5–14.1 µs) or intracellularly expressed mCherry (HEK293 cells, 1.54–1.67 ns), but a decrease in fluorescence lifetime of GFP from 2.65 to 2.29 ns (in V15B cells). These values represent a viscosity change from 8.59 to 20.56 cP as well as significant changes in the refractive index (RI), according to independent calibration. Interestingly DNA damage localized to a submicron region following a laser microbeam induction showed a whole cell viscosity change, with those in the nucleus being greater than the cytoplasm. We also found evidence of a by-stander effect, whereby adjacent un-irradiated cells also showed nuclear viscosity change. Finally, an increase in viscosity following DNA damage was also observed in bacterial cells with an over-expressed mNeonGreen FP, evidenced by the change in its lifetime from 2.8 to 2.4 ns.

Deep learning (DL) frameworks serve as the backbone for a wide range of artificial intelligence applications. However, bugs within DL frameworks can cascade into critical issues in higher-level applications, jeopardizing reliability and security. While numerous techniques have been proposed to detect bugs in DL frameworks, research exploring common API patterns across frameworks and the potential risks they entail remains limited. Notably, many DL frameworks expose similar APIs with overlapping input parameters and functionalities, rendering them vulnerable to shared bugs, where a flaw in one API may extend to analogous APIs in other frameworks. To address this challenge, we propose MirrorFuzz, an automated API fuzzing solution to discover shared bugs in DL frameworks. MirrorFuzz operates in three stages: First, MirrorFuzz collects historical bug data for each API within a DL framework to identify potentially buggy APIs. Second, it matches each buggy API in a specific framework with similar APIs within and across other DL frameworks. Third, it employs large language models (LLMs) to synthesize code for the API under test, leveraging the historical bug data of similar APIs to trigger analogous bugs across APIs. We implement MirrorFuzz and evaluate it on four popular DL frameworks (TensorFlow, PyTorch, OneFlow, and Jittor). Extensive evaluation demonstrates that MirrorFuzz improves code coverage by 39.92% and 98.20% compared to state-of-the-art methods on TensorFlow and PyTorch, respectively. Moreover, MirrorFuzz discovers 315 bugs, 262 of which are newly found, and 80 bugs are fixed, with 52 of these bugs assigned CNVD IDs.

Camptothecin (CPT) is a topoisomerase I inhibitor, derivatives of which are being used for cancer chemotherapy. CPT-induced DNA double-strand breaks (DSBs) are considered a major cause of its tumoricidal activity, and it has been shown that CPT induces DNA damage signaling through the phosphatidylinositol 3-kinase-related kinases, including ATM (ataxia telangiectasia mutated), ATR (ATM and Rad3-related), and DNA-PK (DNA-dependent protein kinase). In addition, CPT causes DNA strand breaks mediated by transcription, although the downstream signaling events are less well characterized. In this study, we show that CPT-induced activation of ATM requires transcription. Mechanistically, transcription inhibition suppressed CPT-dependent activation of ATM and blocked recruitment of the DNA damage mediator p53-binding protein 1 (53BP1) to DNA damage sites, whereas ATM inhibition abrogated CPT-induced G(1)/S and S phase checkpoints. Functional inactivation of ATM resulted in DNA replication-dependent hyperactivation of DNA-PK in CPT-treated cells and dramatic CPT hypersensitivity. On the other hand, simultaneous inhibition of ATM and DNA-PK partially restored CPT resistance, suggesting that activation of DNA-PK is proapoptotic in the absence of ATM. Correspondingly, comet assay and cell cycle synchronization experiments suggested that transcription collapse occurring as the result of CPT treatment are converted to frank double-strand breaks when ATM-deficient cells bypass the G(1)/S checkpoint. Thus, ATM suppresses DNA-PK-dependent cell death in response to topoisomerase poisons, a finding with potential clinical implications.

Educational Data Mining (EDM) plays a key role in improving modern learning by using advanced techniques. This study focuses on predicting college dropout rates with a deep learning framework. By analysing student data, our model helps identify at-risk students early, allowing timely support. The dataset contains student details such as academic performance, socioeconomic status, and engagement levels. Using these details, our framework applies deep learning methods to detect early signs of dropout. This helps universities take quick action to support students and improve their learning experience. Our study checked how well different models predicted dropout rates. The K-Nearest Neighbours (KNN) model had 93% accuracy, while the Recurrent Neural Network (RNN) had 95% accuracy. When KNN and RNN were combined, accuracy dropped to 87%. However, combining KNN and Convolutional Neural Network (CNN) improved accuracy to 97%. The RNN and CNN combination scored 96% accuracy, and the standalone CNN model also reached 97% accuracy. The best-performing method in this study is the KNN and CNN combination, which achieved the highest accuracy of 97%. Even though the CNN model alone also reached 97%, combining it with KNN helps capture more patterns in the data. This approach improves student retention by identifying and helping struggling students early. Our findings show that deep learning can effectively predict dropout rates, allowing educational institutions to provide better student support and improve success rates.

Staphylococcus aureus is a common bacterial etiology of serious infectious diseases. S. aureus can invade various types of non-professional phagocytes to produce host cell death. We show here that shortly after invasion of HeLa cells S. aureus transit to autophagosomes was characterized by double membranes and co-localization with LC3. S. aureus were not able to replicate and produce cell death in autophagy-deficient atg5-/- mouse embryonic fibroblasts. S. aureus-containing autophagosomes do not acidify nor do they acquire lysosome-associated membrane protein-2, indicating that S. aureus inhibits autophagosome maturation and fusion with lysosomes. Eventually, S. aureus escape from autophagosomes into the cytoplasm, which results in caspase-independent host cell death. S. aureus strains deficient for agr, a global regulator of S. aureus virulence, were not targeted by autophagy and did not produce host-cell death. Autophagy induction by rapamycin restored both replication and cytotoxicity of agr-deficient S. aureus strains, indicating that an agr-regulated factor(s) is required for autophagy-mediated cytotoxicity. The results of this study suggest that rapid induction of autophagy is essential for S. aureus replication, escape into the cytoplasm, and host cell killing.