Chirality-Dependent Dynamic Evolution for Trions in Monolayer WS2.

Monolayer transition metal dichalcogenides exhibit valley-dependent excitonic characters with a large binding energy, acting as the building block for future optoelectronic functionalities. Herein, combined with pump-probe ultrafast transient transmission spectroscopy and theoretical simulations, we reveal the chirality-dependent trion dynamics in h-BN encapsulated monolayer tungsten disulfide. By resonantly pumping trions in a single valley and monitoring their temporal evolution, we identify the temperature-dependent competition between two relaxation channels driven by chirality-dependent scattering processes. At room temperature, the phonon-assisted upconversion process predominates, converting excited trions to excitons within the same valley on a sub-picosecond (ps) time scale. As temperature decreases, this process becomes less efficient, while alternative channels, notably valley depolarization process for trions, assume importance, leading to an increase of trion density in the unpumped valley within a ps time scale. Our time-resolved valley-contrast results provide a comprehensive insight into trion dynamics in 2D materials, thereby advancing the development of novel valleytronic devices.

Ferritinophagy mediates adaptive resistance to EGFR tyrosine kinase inhibitors in non-small cell lung cancer.

Osimertinib (Osi) is a widely used epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI). However, the emergence of resistance is inevitable, partly due to the gradual evolution of adaptive resistant cells during initial treatment. Here, we find that Osi treatment rapidly triggers adaptive resistance in tumor cells. Metabolomics analysis reveals a significant enhancement of oxidative phosphorylation (OXPHOS) in Osi adaptive-resistant cells. Mechanically, Osi treatment induces an elevation of NCOA4, a key protein of ferritinophagy, which maintains the synthesis of iron-sulfur cluster (ISC) proteins of electron transport chain and OXPHOS. Additionally, active ISC protein synthesis in adaptive-resistant cells significantly increases the sensitivity to copper ions. Combining Osi with elesclomol, a copper ion ionophore, significantly increases the efficacy of Osi, with no additional toxicity. Altogether, this study reveals the mechanisms of NCOA4-mediated ferritinophagy in Osi adaptive resistance and introduces a promising new therapy of combining copper ionophores to improve its initial efficacy.

Construction and verification of a histone deacetylases-related prognostic signature model for colon cancer.

Histone deacetylases (HDACs) contribute significantly to the initiation, progression, and prognosis of colorectal adenocarcinoma (COAD). Additionally, HDACs regulate the tumor microenvironment, immune escape, and tumor stem cells, and are closely linked to COAD prognosis. We developed a prognostic model for COAD that incorporates HDACs to evaluate their specific roles. The COAD dataset containing clinical and mutation data was collected using the TCGA and GEO databases to obtain genes associated with HDAC. LASSO analysis and univariate and multivariate Cox regression analysis were used to determine the presence of prognostic genes. Multivariate Cox analysis was also used to determine risk scores for HDAC-related features. Furthermore, genomic alterations, immune infiltration, and drug response were compared between high- and low-risk groups. Cellular experiments validated the potential regulatory role of BRD3 on COAD proliferation, migration, and apoptosis. The median risk scores, calculated based on the characteristics, demonstrated a more significant prognostic improvement in patients in the low-risk group. Furthermore, HDAC-related features were identified as important independent prognostic factors for patients with COAD. Additionally, genomic mutation status, immune infiltration, and function, as well as response to immunotherapy and chemotherapy, were found to be associated with risk scores. Subgroup analyses indicate that anti-PD-1 therapy may be beneficial for patients in the low-risk group. Additionally, a decrease in risk score was associated with a decrease in immune infiltration. Finally, HCT116 and HT29 cells exhibited inhibition of BRD3 gene proliferation and migration, as well as promotion of apoptosis. In patients with COAD, HDAC-related characteristics may be useful in predicting survival and selecting treatment.

Exciton polariton condensation from bound states in the continuum at room temperature.

Exciton-polaritons (polaritons) resulting from the strong exciton-photon interaction stimulates the development of novel low-threshold coherent light sources to circumvent the ever-increasing energy demands of optical communications1-3. Polaritons from bound states in the continuum (BICs) are promising for Bose-Einstein condensation owing to their theoretically infinite quality factors, which provide prolonged lifetimes and benefit the polariton accumulations4-7. However, BIC polariton condensation remains limited to cryogenic temperatures ascribed to the small exciton binding energies of conventional material platforms. Herein, we demonstrated room-temperature BIC polariton condensation in perovskite photonic crystal lattices. BIC polariton condensation was demonstrated at the vicinity of the saddle point of polariton dispersion that generates directional vortex beam emission with long-range coherence. We also explore the peculiar switching effect among the miniaturized BIC polariton modes through effective polariton-polariton scattering. Our work paves the way for the practical implementation of BIC polariton condensates for integrated photonic and topological circuits.

Energy transfer driven brightening of MoS2 by ultrafast polariton relaxation in microcavity MoS2/hBN/WS2 heterostructures.

Energy transfer is a ubiquitous phenomenon that delivers energy from a blue-shifted emitter to a red-shifted absorber, facilitating wide photonic applications. Two-dimensional (2D) semiconductors provide unique opportunities for exploring novel energy transfer mechanisms in the atomic-scale limit. Herein, we have designed a planar optical microcavity-confined MoS2/hBN/WS2 heterojunction, which realizes the strong coupling among donor exciton, acceptor exciton, and cavity photon mode. This configuration demonstrates an unconventional energy transfer via polariton relaxation, brightening MoS2 with a record-high enhancement factor of ~440, i.e., two-order-of-magnitude higher than the data reported to date. The polariton relaxation features a short characteristic time of ~1.3 ps, resulting from the significantly enhanced intra- and inter-branch exciton-exciton scattering. The polariton relaxation dynamics is associated with Rabi energies in a phase diagram by combining experimental and theoretical results. This study opens a new direction of microcavity 2D semiconductor heterojunctions for high-brightness polaritonic light sources and ultrafast polariton carrier dynamics.

Development and validation of a semi-supervised deep learning model for automatic retinopathy of prematurity staging.

Retinopathy of prematurity (ROP) is currently one of the leading causes of infant blindness worldwide. Recently significant progress has been made in deep learning-based computer-aided diagnostic methods. However, deep learning often requires a large amount of annotated data for model optimization, but this requires long hours of effort by experienced doctors in clinical scenarios. In contrast, a large number of unlabeled images are relatively easy to obtain. In this paper, we propose a new semi-supervised learning framework to reduce annotation costs for automatic ROP staging. We design two consistency regularization strategies, prediction consistency loss and semantic structure consistency loss, which can help the model mine useful discriminative information from unlabeled data, thus improving the generalization performance of the classification model. Extensive experiments on a real clinical dataset show that the proposed method promises to greatly reduce the labeling requirements in clinical scenarios while achieving good classification performance.

Brain Structural Connectivity Guided Vision Transformers for Identification of Functional Connectivity Characteristics in Preterm Neonates.

Preterm birth is the leading cause of death in children under five years old, and is associated with a wide sequence of complications in both short and long term. In view of rapid neurodevelopment during the neonatal period, preterm neonates may exhibit considerable functional alterations compared to term ones. However, the identified functional alterations in previous studies merely achieve moderate classification performance, while more accurate functional characteristics with satisfying discrimination ability for better diagnosis and therapeutic treatment is underexplored. To address this problem, we propose a novel brain structural connectivity (SC) guided Vision Transformer (SCG-ViT) to identify functional connectivity (FC) differences among three neonatal groups: preterm, preterm with early postnatal experience, and term. Particularly, inspired by the neuroscience-derived information, a novel patch token of SC/FC matrix is defined, and the SC matrix is then adopted as an effective mask into the ViT model to screen out input FC patch embeddings with weaker SC, and to focus on stronger ones for better classification and identification of FC differences among the three groups. The experimental results on multi-modal MRI data of 437 neonatal brains from publicly released Developing Human Connectome Project (dHCP) demonstrate that SCG-ViT achieves superior classification ability compared to baseline models, and successfully identifies holistically different FC patterns among the three groups. Moreover, these different FCs are significantly correlated with the differential gene expressions of the three groups. In summary, SCG-ViT provides a powerfully brain-guided pipeline of adopting large-scale and data-intensive deep learning models for medical imaging-based diagnosis.

A Unified and Biologically Plausible Relational Graph Representation of Vision Transformers.

Vision transformer (ViT) and its variants have achieved remarkable success in various tasks. The key characteristic of these ViT models is to adopt different aggregation strategies of spatial patch information within the artificial neural networks (ANNs). However, there is still a key lack of unified representation of different ViT architectures for systematic understanding and assessment of model representation performance. Moreover, how those well-performing ViT ANNs are similar to real biological neural networks (BNNs) is largely unexplored. To answer these fundamental questions, we, for the first time, propose a unified and biologically plausible relational graph representation of ViT models. Specifically, the proposed relational graph representation consists of two key subgraphs: an aggregation graph and an affine graph. The former considers ViT tokens as nodes and describes their spatial interaction, while the latter regards network channels as nodes and reflects the information communication between channels. Using this unified relational graph representation, we found that: 1) model performance was closely related to graph measures; 2) the proposed relational graph representation of ViT has high similarity with real BNNs; and 3) there was a further improvement in model performance when training with a superior model to constrain the aggregation graph.

Application of artificial intelligence system for screening multiple fundus diseases in Chinese primary healthcare settings: a real-world, multicentre and cross-sectional study of 4795 cases.

BACKGROUND/AIMS: This study evaluates the performance of the Airdoc retinal artificial intelligence system (ARAS) for detecting multiple fundus diseases in real-world scenarios in primary healthcare settings and investigates the fundus disease spectrum based on ARAS. METHODS: This real-world, multicentre, cross-sectional study was conducted in Shanghai and Xinjiang, China. Six primary healthcare settings were included in this study. Colour fundus photographs were taken and graded by ARAS and retinal specialists. The performance of ARAS is described by its accuracy, sensitivity, specificity and positive and negative predictive values. The spectrum of fundus diseases in primary healthcare settings has also been investigated. RESULTS: A total of 4795 participants were included. The median age was 57.0 (IQR 39.0-66.0) years, and 3175 (66.2%) participants were female. The accuracy, specificity and negative predictive value of ARAS for detecting normal fundus and 14 retinal abnormalities were high, whereas the sensitivity and positive predictive value varied in detecting different abnormalities. The proportion of retinal drusen, pathological myopia and glaucomatous optic neuropathy was significantly higher in Shanghai than in Xinjiang. Moreover, the percentages of referable diabetic retinopathy, retinal vein occlusion and macular oedema in middle-aged and elderly people in Xinjiang were significantly higher than in Shanghai. CONCLUSION: This study demonstrated the dependability of ARAS for detecting multiple retinal diseases in primary healthcare settings. Implementing the AI-assisted fundus disease screening system in primary healthcare settings might be beneficial in reducing regional disparities in medical resources. However, the ARAS algorithm must be improved to achieve better performance. TRIAL REGISTRATION NUMBER: NCT04592068.

ASSOCIATION OF TESSELLATION DENSITY WITH PROGRESSION OF AXIAL LENGTH AND REFRACTION IN CHILDREN: An Artificial Intelligence-Assisted 4-Year Study.

PURPOSE: To investigate fundus tessellation density (TD) and its association with axial length (AL) elongation and spherical equivalent (SE) progression in children. METHODS: The school-based prospective cohort study enrolled 1,997 individuals aged 7 to 9 years in 11 elementary schools in Mojiang, China. Cycloplegic refraction and biometry were performed at baseline and 4-year visits. The baseline fundus photographs were taken, and TD, defined as the percentage of exposed choroidal vessel area in the photographs, was quantified using an artificial intelligence-assisted semiautomatic labeling approach. After the exclusion of 330 ineligible participants because of loss to follow-up or ineligible fundus photographs, logistic models were used to assess the association of TD with rapid AL elongation (>0.36 mm/year) and SE progression (>1.00 D/year). RESULTS: The prevalence of tessellation was 477 of 1,667 (28.6%) and mean TD was 0.008 ± 0.019. The mean AL elongation and SE progression in 4 years were 0.90 ± 0.58 mm and -1.09 ± 1.25 D. Higher TD was associated with longer baseline AL (ß, 0.030; 95% confidence interval: 0.015-0.046; P < 0.001) and more myopic baseline SE (ß, -0.017; 95% confidence interval: -0.032 to -0.002; P = 0.029). Higher TD was associated with rapid AL elongation (odds ratio, 1.128; 95% confidence interval: 1.055-1.207; P < 0.001) and SE progression (odds ratio, 1.123; 95% confidence interval: 1.020-1.237; P = 0.018). CONCLUSION: Tessellation density is a potential indicator of rapid AL elongation and refractive progression in children. TD measurement could be a routine to monitor AL elongation.

Effect of vault on predicting postoperative refractive error for posterior chamber phakic intraocular lens based on a machine learning model.

PURPOSE: To investigate how vault and other biometric variations affect postoperative refractive error of implantable collamer lenses (ICLs) by integrating artificial intelligence and modified vergence formula. SETTING: Eye and ENT Hospital of Fudan University, Shanghai, China. DESIGN: Artificial intelligence and big data-based prediction model. METHODS: 2845 eyes that underwent uneventful spherical ICL or toric ICL implantation and with manifest refraction results 1 month postoperatively were included. 1 eye of each patient was randomly included. Random forest was used to calculate the postoperative sphere, cylinder, and spherical equivalent by inputting variable ocular parameters. The influence of predicted vault and modified Holladay formula on predicting postoperative refractive error was analyzed. Subgroup analysis of ideal vault (0.25 to 0.75 mm) and extreme vault (<0.25 mm or >0.75 mm) was performed. RESULTS: In the test set of both ICLs, all the random forest-based models significantly improved the accuracy of predicting postoperative sphere compared with the Online Calculation & Ordering System calculator ( P < .001). For ideal vault, the combination of modified Holladay formula in spherical ICL exhibited highest accuracy ( R = 0.606). For extreme vault, the combination of predicted vault in spherical ICL enhanced R values ( R = 0.864). The combination of predicted vault and modified Holladay formula was most optimal for toric ICL in all ranges of vault (ideal vault: R = 0.516, extreme vault: R = 0.334). CONCLUSIONS: The random forest-based calculator, considering vault and variable ocular parameters, illustrated superiority over the existing calculator on the study datasets. Choosing an appropriate lens size to control the vault within the ideal range was helpful to avoid refractive surprises.

Analysis of the effect of Ivor-Lewis esophagectomy and McKeown esophagectomy on perioperative anxiety and depression in patients with esophageal cancer.

To compare the effects of Ivor-Lewis esophagectomy and McKeown esophagectomy on perioperative anxiety and depression in patients with esophageal cancer. Sixty-three patients with stage I-III middle and lower esophageal carcinoma from June 2021 to December 2022 were randomly divided into observation group (n = 32) treated with laparoscopic Ivor-Lewis esophagectomy and control group (n = 31) treated with laparoscopic McKeown esophagectomy. Self-Rating Depression Scale (SDS) and Self-Rating Anxiety Scale (SAS) were measured on the second day of admission and the fifth day after surgery to assess the presence of depression and anxiety. The preoperative and postoperative clinical data of both groups were compared, and multivariate analysis was used to identify risk factors associated with depression and anxiety in patients with esophageal cancer. There was no significant difference in SDS and SAS standard scores between the observation group and the control group ( P  > 0.05). The postoperative SDS and SAS scores in the control group were significantly higher than those before and after operation in the observation group ( P  < 0.01). According to univariate analysis, patients with TNM stage III, tumor diameter greater than 3 cm, postoperative complications, radical McKeown esophagectomy, and C-reactive protein levels above 10 mg/L had a higher incidence of depression and anxiety ( P  < 0.05). Multivariate logistic analysis showed that TNM stage III (depression: OR 1.683, 95 CI 1.429-1.861; Anxiety: OR 1.739, 95 CI 1.516-1.902), postoperative complications (depression: OR 2.345, 95 CI 1.435-3.891; Anxiety: OR 1.872, 95 CI 1.372-3.471), surgical approach (depression: OR 1.609, 95 CI 1.502-3.193; Anxiety: OR 1.658, 95 CI 1.469-2.059), and C-reactive protein (depression: OR 2.260, 95 CI 1.157-4.059; Anxiety: OR 0.373, 95 CI 0.253-0.976) were all independent factors for depression and anxiety in patients after esophageal cancer surgery ( P  < 0.05). The Ivor-Lewis esophagectomy has the advantages of fewer complications and low inflammatory response, which can help alleviate anxiety and depression and improve patients' quality of life and prognosis.

Regularity and variability of functional brain connectivity characteristics between gyri and sulci under naturalistic stimulus.

The human cerebral cortex is folded into two fundamentally anatomical units: gyri and sulci. Previous studies have demonstrated the genetical, structural, and functional differences between gyri and sulci, providing a unique perspective for revealing the relationship among brain function, cognition, and behavior. While previous studies mainly focus on the functional differences between gyri and sulci under resting or task-evoked state, such characteristics under naturalistic stimulus (NS) which reflects real-world dynamic environments are largely unknown. To address this question, this study systematically investigates spatio-temporal functional connectivity (FC) characteristics between gyri and sulci under NS using a spatio-temporal graph convolutional network model. Based on the public Human Connectome Project dataset of 174 subjects with four different runs of both movie-watching NS and resting state 7T functional MRI data, we successfully identify unique FC features under NS, which are mainly involved in visual, auditory, emotional and cognitive control, and achieve high discriminative accuracy 93.06 % to resting state. Moreover, gyral regions as well as gyro-gyral connections consistently participate more as functional information exchange hubs than sulcal ones among these networks. This study provides novel insights into the functional brain mechanism under NS and lays a solid foundation for accurately mapping the brain anatomy-function relationship.

Different associations between organ-specific immune-related adverse event and survival in non-small cell lung cancer patients treated with programmed death-1 inhibitors-based combination therapy.

Background: The profile of immune-related adverse events (irAEs) due to programmed death-1 (PD-1) inhibitors-based combination therapy in advanced non-small cell lung cancer (NSCLC) and its relationship with survival have not been fully described. Objective: Designed to capture the spectrum of irAEs and explore the association between irAEs and clinical outcomes in patients with NSCLC. Design: This retrospective single-center study included patients with advanced NSCLC treated with PD-1 inhibitors (mainly in combination with chemotherapy) at Jiangsu Cancer Hospital. Methods: The relationship between irAEs and survival was explored using landmark analysis and time-dependent Cox regression. The subgroup analyses focused on investigating the effects of organ-specific irAE, irAE grade, and steroid dose used to treat irAE. Results: This study included 301 patients, 199 of whom received PD-1 inhibitors plus chemotherapy. The most common irAEs were skin toxicity (19.3%), endocrinopathy (21.3%), and pneumonitis (17.6%). In the entire cohort, the median progression-free survival (PFS) for patients developing and not developing irAE was 12.3 and 10.7 months (p < 0.001), and the median overall survival (OS) was 23.5 months and 20.1 months (p = 0.137), respectively. Subgroup analyses indicated that grade 3 or higher irAE, high steroid dose, and immune-related pneumonitis were detrimental to OS, whereas skin toxicity was beneficial to survival. These findings were further corroborated by both landmark analyses and Cox regression models conducted over four time points (1, 3, 6, and 12 months). Conclusion: In the real world, NSCLC patients receiving PD-1 inhibitor-based combination therapy (particularly combined with chemotherapy) experience longer PFS with irAE, though not necessarily OS. Immune-related skin toxicity is associated with a better prognosis, whereas pneumonitis grade ⩾3 irAE and high steroid dose compromise survival. Clinicians should remain cognizant of the organ-specific manifestations of irAE and take proactive measures to mitigate the progression of irAE.

Integrating Network Pharmacology and Experimental Validation to Decipher the Anti-Inflammatory Effects of Magnolol on LPS-induced RAW264.7 Cells.

INTRODUCTION: Magnolol is beneficial against inflammation-mediated damage. However, the underlying mechanisms by which m+agnolol exerts anti-inflammatory effects on macrophages remain unclear. OBJECTIVE: In this study, network pharmacology and experimental validation were used to assess the effect of magnolol on inflammation caused by lipopolysaccharide (LPS) in RAW264.7 cells. MATERIALS AND METHODS: Genes related to magnolol were identified in the PubChem and Swiss Target Prediction databases, and gene information about macrophage polarization was retrieved from the GeneCards, OMIM, and PharmGKB databases. Analysis of protein-protein interactions was performed with STRING, and Cytoscape was used to construct a component-target-disease network. GO and KEGG enrichment analyses were performed to ascertain significant molecular biological processes and signaling pathways. LPS was used to construct the inflammatory cell model. ELISA and qRT‒PCR were used to examine the expression levels of inflammationassociated factors, immunofluorescence was used to examine macrophage markers (CD86 and CD206), and western blotting was used to examine protein expression levels. RESULTS: The hub target genes of magnolol that act on macrophage polarization were MDM2, MMP9, IL-6, TNF, EGFR, AKT1, and ERBB2. The experimental validation results showed that magnolol treatment decreased the levels of proinflammatory factors (TNF-α, IL-1ß, and IL-6). Moreover, the levels of anti-inflammatory factors (IL-10 and IL-4) were increased. In addition, magnolol upregulated the expression of M2 markers (Agr-1, Fizzl, and CD206) and downregulated M1 markers (CD86). The cell experiment results supported the network pharmacological results and demonstrated that magnolol alleviated inflammation by modulating the PI3k-Akt and P62/keap1/Nrf2 signaling pathways. CONCLUSION: According to network pharmacology and experimental validation, magnolol attenuated inflammation in LPS-induced RAW264.7 cells mainly by inhibiting M1 polarization and enhancing M2 polarization by activating the PI3K/Akt and P62/keap1/Nrf2 signaling pathways.

N6 -methyladenosine-modified circFUT8 competitively interacts with YTHDF2 and miR-186-5p to stabilize FUT8 mRNA to promote malignant progression in lung adenocarcinoma.

BACKGROUND: Lung cancer is the leading cause of cancer related to mortality worldwide, and the main pathological type is lung adenocarcinoma (LUAD). Circular RNAs (circRNAs) have been reported to be modified by N6 -methyladenosine (m6A), which is involved in the progression of diverse tumors. However, the crosstalk between circRNAs and m6A modification has not been well elucidated in LUAD. METHODS: MeRIP-seq and YTHDF2-RIP-seq datasets were explored to identify candidate circRNAs modified by YTHDF2. Dual-luciferase reporter assay, RIP, and rescue assays were performed to explore the relationship between circFUT8 and its parent mRNA of FUT8. In vitro and in vivo experiments were utilized to uncover the function of circFUT8. RESULTS: In this study, we identified a novel m6A-modified circFUT8, derived from exon 3 of FUT8, which was elevated in tumor tissues compared with adjacent noncancerous tissues. The m6A reader YTHDF2 recognized and destabilized circFUT8 in an m6A-dependent manner. YTHDF2 also combined with the line form of FUT8 (mFUT8), and circFUT8 competitively interacted with YTHDF2, blunting its binding to mFUT8, to stabilize the mRNA level of FUT8. Additionally, circFUT8 sponged miR-186-5p to elevate the expression of mFUT8. Finally, we revealed that circFUT8 promoted the malignant progression of LUAD dependent on the oncogenic function of FUT8. CONCLUSIONS: These findings identified a novel m6A-modified circFUT8 recognized and destabilized by YTHDF2, which competitively interacted with YTHDF2 and miR-186-5p to stabilize FUT8 mRNA to promote malignant progression in LUAD.

C-IGF1R encoded by cIGF1R acts as a molecular switch to restrict mitophagy of drug-tolerant persister tumour cells in non-small cell lung cancer.

The clinical efficacy of Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitors (EGFR-TKIs) is limited by the emergence of drug resistance. We hypothesise that restoring dysregulated circular RNAs under initial treatment with EGFR-TKIs may enhance their effectiveness. Through high-throughput screening, we identify that combining circular RNA IGF1R (cIGF1R) with EGFR-TKIs significantly synergises to suppress tumour regrowth following drug withdrawal. Mechanistically, cIGF1R interacts with RNA helicase A (RHA) to depress insulin-like growth factor 1 receptor (IGF1R) mRNA splicing, negatively regulating the parent IGF1R signalling pathway. This regulation is similar to that of IGF1R inhibitor, which induces drug-tolerant persister (DTP) state with activated mitophagy. The cIGF1R also encodes a peptide C-IGF1R that reduces Parkin-mediated ubiquitination of voltage-dependent anion channel 1 (VDAC1) to restrict mitophagy, acting as a molecular switch that promotes the transition of DTP to apoptosis. Our study shows that combining cIGF1R with EGFR-TKIs efficiently reduces the emergence of DTP.

Unraveling the Ultrafast Coherent Dynamics of Exciton Polariton Propagation at Room Temperature.

Exciton polaritons are widely considered as promising platforms for developing room-temperature polaritonic devices, owing to the high-speed propagation and nonlinear interactions. However, it remains challenging to explore the dynamics of exciton polaritons specifically at room temperature, where the lifetime could be as small as a few picoseconds and the prevailing time-averaged measurement cannot give access to the true nature of it. Herein, by using the time-resolved photoluminescence, we have successfully traced the ultrafast coherent dynamics of a moving exciton polariton condensate in a one-dimensional perovskite microcavity. The propagation speed is directly measured to be ∼12.2 ± 0.8 µm/ps. Moreover, we have developed a time-resolved Michelson interferometry to quantify the time-dependent phase coherence, which reveals that the actual coherence time of exciton polaritons could be much longer (nearly 100%) than what was believed before. Our work sheds new light on the ultrafast coherent propagation of exciton polaritons at room temperature.

Rectify ViT Shortcut Learning by Visual Saliency.

Shortcut learning in deep learning models occurs when unintended features are prioritized, resulting in degenerated feature representations and reduced generalizability and interpretability. However, shortcut learning in the widely used vision transformer (ViT) framework is largely unknown. Meanwhile, introducing domain-specific knowledge is a major approach to rectifying the shortcuts that are predominated by background-related factors. For example, eye-gaze data from radiologists are effective human visual prior knowledge that has the great potential to guide the deep learning models to focus on meaningful foreground regions. However, obtaining eye-gaze data can still sometimes be time-consuming, labor-intensive, and even impractical. In this work, we propose a novel and effective saliency-guided ViT (SGT) model to rectify shortcut learning in ViT with the absence of eye-gaze data. Specifically, a computational visual saliency model (either pretrained or fine-tuned) is adopted to predict saliency maps for input image samples. Then, the saliency maps are used to filter the most informative image patches. Considering that this filter operation may lead to global information loss, we further introduce a residual connection that calculates the self-attention across all the image patches. The experiment results on natural and medical image datasets show that our SGT framework can effectively learn and leverage human prior knowledge without eye-gaze data and achieves much better performance than baselines. Meanwhile, it successfully rectifies the harmful shortcut learning and significantly improves the interpretability of the ViT model, demonstrating the promise of transferring human prior knowledge derived visual saliency in rectifying shortcut learning.

The super-enhancer-driven lncRNA LINC00880 acts as a scaffold between CDK1 and PRDX1 to sustain the malignance of lung adenocarcinoma.

Super-enhancers (SEs) are regulatory element clusters related to cell identity and disease. While the studies illustrating the function of SE-associated long noncoding RNAs (lncRNAs) in lung adenocarcinoma (LUAD) remains few. In our research, a SE-driven lncRNA, LINC00880, was identified, which showed higher expression in LUAD compared to normal tissues and indicated worse outcomes in stage I LUADs. We found that the transcription factor (TF) FOXP3 could simultaneously occupy the promoter and SE regions of LINC00880 to promote its transcription. The oncogenic function of LINC00880 was validated both in vitro and in vivo. Mechanically, LINC00880 binds to the protein CDK1 to increase its kinase activity, which rely on the phosphorylation state of pT161 in CDK1. LINC00880 also promotes the interaction between CDK1 and PRDX1. Moreover, LINC00880 interacts with PRDX1, which indicates that LINC00880 acts as a protein scaffold between CDK1 and PRDX1 to form a ternary complex, thereby resulting in the activation of PI3K/AKT to promote malignancy. Our results reveal that the SE-associated lncRNA LINC00880 regulates the CDK1/PRDX1 axis to sustain the malignancy of LUAD, providing a novel therapeutic target.