The rapid growth in dataset sizes in modern deep learning has significantly increased data storage costs. Furthermore, the training and time costs for deep neural networks are generally proportional ...to the dataset size. Therefore, reducing the dataset size while maintaining model performance is an urgent research problem that needs to be addressed. Dataset condensation is a technique that aims to distill the original dataset into a much smaller synthetic dataset while maintaining downstream training performance on any agnostic neural network. Previous work has demonstrated that matching the training trajectory between the synthetic dataset and the original dataset is more effective than matching the instantaneous gradient, as it incorporates long-range information. Despite the effectiveness of trajectory matching, it suffers from complex gradient unrolling across iterations, which leads to significant memory and computation overhead. To address this issue, this paper proposes a novel approach called Expert Subspace Projection (ESP), which leverages long-range information while avoiding gradient unrolling. Instead of strictly enforcing the synthetic dataset’s training trajectory to mimic that of the real dataset, ESP only constrains it to lie within the subspace spanned by the training trajectory of the real dataset. The memory-saving advantage offered by our method facilitates unbiased training on the complete set of synthetic images and seamless integration with other dataset condensation techniques. Through extensive experiments, we have demonstrated the effectiveness of our approach. Our method outperforms the trajectory matching method on CIFAR10 by 16.7% in the setting of 1 Image/Class, surpassing the previous state-of-the-art method by 3.2%.
Prodigiosin (PG), a member of a family of natural red pigments produced by a variety of bacteria, was first discovered in Serratia marcescens. PG has been reported to have an apoptosis-inducing ...effect in many cancers, such as lymphoma, colon cancer and nasopharyngeal carcinoma. For this study, we used three glioblastoma (GBM) cell lines (LN229, U251 and A172) to explore the effect of prodigiosin on GBM cells. A CCK8 assay was used to evaluate cell viability. We determinedthe cell cycle distribution by flow cytometry and measured proliferation by an EdU incorporation assay. The expression of different molecules was investigated by western blotting and RT-PCR. We further confirmed our results by plasmid transfection and lentiviral transduction. The LN229 xenograft model was used to study the effect of prodigiosin in vivo. We confirmed that prodigiosin played an anticancer role in several GBM cell lines through the KIAA1524/PP2A/Akt signalling pathway. Prodigiosin inhibited the protein expression of KIAA1524 by suppressing its transcription, which led to activation of PP2A. Afterward, PP2A inhibited the phosphorylation of Akt, thereby inducing increased expression of p53/p21. Furthermore, it was verified that prodigiosin inhibited the KIAA1524/PP2A/Akt axis in vivo in the LN229 xenograft model. These data improve the understanding of the anticancer effects of prodigiosin and further highlight the potential of prodigiosin for the development of anti-glioma drugs.
2,5-Dimethyl-celecoxib (DMC) is a close structural analog of the selective COX-2 inhibitor celecoxib that lacks COX-2-inhibitory function. Thus, DMC is a promising drug for anti-tumor. In this study, ...we evaluated the efficacy and the molecular basis of DMC in the treatment of human glioblastoma multiforme (GBM). DMC inhibited the growth and proliferation of GBM cell lines (LN229, A172, U251, and U87MG) in a dose-dependent manner (
P
< 0.001). In GBM cells treated with DMC, detection by flow cytometry showed cell cycle arrest, and proteins involved in cell cycle such as P21 were increased. Compared with control group, Annexin-V/PI-staining in DMC-treatment group was increased, indicating that DMC could induce apoptosis in GBM cells. Also, associated proteins including cleaved caspase 3 and cleaved PARP-1 were increased. It was further explored whether DMC blocked cell cycle and induced apoptosis in GBM cells through CIP2A/PP2A/AKT signaling pathway. After treatment of DMC, the phosphorylation of Akt was reduced while the total Akt level was not affected. DMC suppressed the expression of CIP2A in a time-dependent manner, while the CIP2A overexpression group reversed cell cycle and apoptotic protein expression led by DMC. Finally, in a xenograft model in nude mice using LN229 cells, DMC suppressed tumor growth. These findings proved that DMC could block cell cycle and induce apoptosis in GBM cells by suppressing CIP2A/PP2A/Akt signaling axis, which indicated that DMC could be an effective option for GBM treatment.
Display omitted
Thermoelectric materials have a wide range of application because they can be directly used in refrigeration and power generation. And the Bi2Te3 stand out because of its excellent ...thermoelectric performance and are used in commercial thermoelectric devices. However, n-type Bi2Te3 has seriously hindered the development of Bi2Te3-based thermoelectric devices due to its weak mechanical properties and inferior thermoelectric performance. Therefore, it is urgent to develop a high-performance n-type Bi2Te3 polycrystalline. In this work, we employed interstitial Cu and the hot deformation process to optimize the thermoelectric properties of Bi2Te2.7Se0.3, and a high-performance thermoelectric module was fabricated based on this material. Our combined theoretical and experimental effort indicates that the interstitial Cu reduce the defect density in the matrix and suppresses the donor-like effect, leading to a lattice plainification effect in the material. In addition, the two-step hot deformation process significantly improves the preferred orientation of the material and boosts the mobility. As a result, a maximum ZT of 1.27 at 373 K and a remarkable high ZTave of 1.22 across the temperature range of 300–425 K are obtained. The thermoelectric generator (TEG, 7-pair) and thermoelectric cooling (TEC, 127-pair) modules were fabricated with our n-type textured Cu0.01Bi2Te2.7Se0.3 coupled with commercial p-type Bi2Te3. The TEC module demonstrates superior cooling efficiency compared with the commercial Bi2Te3 device, achieving a ΔT of 65 and 83.4 K when the hot end temperature at 300 and 350 K, respectively. In addition, the TEG module attains an impressive conversion efficiency of 6.5% at a ΔT of 225 K, which is almost the highest value among the reported Bi2Te3-based TEG modules.
In this paper, high thermoelectric performance and good shape-conformability were simultaneously obtained in a fine-tuned Ag2+xS0.7Te0.3 compound. The thermoelectric performance of Ag2S0.7Te0.3 was ...enhanced by reducing the Ag content. The trace Ag vacancy plays two main roles in improving thermoelectric performance: first, it lowers the carrier concentration to an optimum range while reserving a high carrier mobility, leads to a high power factor of ∼7.4 μW cm−1 K−2 at 550K; Second, it introduces lattice defects, contributes a lower thermal conductivity. Meanwhile, Ag2+xS0.7Te0.3 samples still possesses good plasticity and processing properties. This study provides a simple yet effective approach to enhance the thermoelectric performance for plastic inorganic thermoelectric materials.
Display omitted
•Diminishing Ag efficiently reduces carrier scattering and increases Seebeck coefficient.•The peak ZT of ∼0.85 at 550 K and the ZTave of ∼0.7 were obtained in Ag1.9992S0.7Te0.3 sample.•Ag1.9992S0.7Te0.3 performs well mechanical properties and have good plasticity and processing properties.
Despite extensive discussions about Hong Kong’s sentiments towards the Chinese mainland since the sovereignty transfer in 1997, there has been a lack of large-scale empirical analysis of attitude ...change in the mainstream media, which both reflect and shape sentiments in society. To address this gap, the present study employs an optimised semantic-based automatic sentiment analysis method to examine a corpus of news about China from 1997 to 2020 in three leading Chinese-language newspapers in Hong Kong, namely Apple Daily, Ming Pao, and Oriental Daily News. The analysis reveals that although the Hong Kong press had a positive emotional tone towards China in general, the overall sentiment was becoming increasingly negative. Meanwhile, the alignment and antagonism towards China have both increased, providing empirical evidence for attitudinal polarisation in the Hong Kong press. Specifically, Apple Daily’s depictions of China became increasingly negative, though with some positive turns before 2008, whereas Oriental Daily News consistently expressed favourable sentiments. Ming Pao maintained an impartial stance towards China through an increased but balanced representation of positive and negative sentiments, with its subjectivity and sentiment intensity growing to an industry-standard level. By explicating the cross-newspaper and cross-period variations, the results provide new insights into the complexity of sentiments towards China in the Hong Kong press, and media attitudes in general in terms of the “us” and “them” positionings.
The 5 September 2022 Mw 6.7 Luding earthquake occurred on the Moxi segment of the highly active Xianshuihe fault in eastern Tibet. Here, we constrain the coseismic slip by jointly inverting the ...coseismic displacements measured by Global Positioning System, seismometer and Interferometric Synthetic Aperture Radar. Along the Moxi fault, concentrated left‐lateral strike slip extends ∼30 km along the strike above 10 km depth, producing 0.7–1.0 m shallow slip. Clustered aftershocks and slip inversions suggest that the secondary conjugate Mozigou fault may also involve the rupture. Southward rupture propagation is likely arrested by the barrier‐like fault segments of the Xianshuihe fault, characterized by high interseismic coupling (>0.6) and reduced shear stress rate (<1 kPa/yr) due to interactions with surrounding large locked asperities. The distribution of aftershocks is highly correlated with the positive coseismic Coulomb failure stress changes , which bring the adjacent asperities on the Anninghe and Daliangshan faults ∼0.2 MPa closer to failure.
Plain Language Summary
The Mw 6.7 Luding earthquake occurred in eastern Tibet on 5 September 2022, causing 93 casualties and 25 people missing, and substantial damage in the epicentral regions. In this study, we derive the static coseismic displacements from Global Positioning System, seismometer and Interferometric Synthetic Aperture Radar observations. We jointly invert these datasets for the coseismic slip distribution. Our results reveal that the coseismic slip is likely distributed on the conjugate Moxi and Mozigou faults, with the primary slip concentrated on the Moxi fault. We compare the coseismic slip with both the interseismic coupling distribution and shear stress rate on the Moxi fault. We find that the southward coseismic rupture termination is spatially coincident with the areas undergoing a low pre‐earthquake shear stress rate and a high interseismic coupling. The comparison suggests that the coseismic slip during the Luding earthquake is likely controlled by pre‐earthquake interactions with surrounding locked zones on the Anninghe and Daliangshan faults (DLS). We calculate the coseismic Coulomb failure stress changes and demonstrate that the aftershocks could be explained by the induced positive stress perturbations. We conclude that the Luding mainshock enhances earthquake hazards on the Anninghe and DLS.
Key Points
Global Positioning System, seismometer, and Interferometric Synthetic Aperture Radar measurements reveal concentrated coseismic slip extending 30 km on the Moxi fault at 0–10 km depth
Interactions with large locked asperities lower the shear stress rate to the south of earthquake and lead to the arrest of coseismic rupture
The Luding event brought the Anninghe and Daliangshan faults ∼0.2 MPa closer to failure, representing enhanced earthquake hazard
A carbon-based solid acid catalyst was prepared by the sulfonation of carbonized vegetable oil asphalt. This catalyst was employed to simultaneously catalyze esterification and transesterification to ...synthesis biodiesel when a waste vegetable oil with large amounts of free fatty acids (FFAs) was used as feedstock. The physical and chemical properties of this catalyst were characterized by a variety of techniques. The maximum conversion of triglyceride and FFA reached 80.5
wt.% and 94.8
wt.% after 4.5
h at 220
°C, when using a 16.8
M ratio of methanol to oil and 0.2
wt.% of catalyst to oil. The high catalytic activity and stability of this catalyst was related to its high acid site density (–OH, Brönsted acid sites), hydrophobicity that prevented the hydration of –OH species, hydrophilic functional groups (–SO
3H) that gave improved accessibility of methanol to the triglyceride and FFAs, and large pores that provided more acid sites for the reactants.
•A DL-based harmonization framework was established with a traveling subject dataset.•Site and brain factors were learned by the proposed framework from gray matter volumes.•Better harmonization ...performance was achieved relative to that of statistics-based methods.•The proposed harmonization method offered flexible expandability for adding new sites.
The accumulation of multisite large-sample MRI datasets collected during large brain research projects in the last decade has provided critical resources for understanding the neurobiological mechanisms underlying cognitive functions and brain disorders. However, the significant site effects observed in imaging data and their derived structural and functional features have prevented the derivation of consistent findings across multiple studies. The development of harmonization methods that can effectively eliminate complex site effects while maintaining biological characteristics in neuroimaging data has become a vital and urgent requirement for multisite imaging studies. Here, we propose a deep learning-based framework to harmonize imaging data obtained from pairs of sites, in which site factors and brain features can be disentangled and encoded. We trained the proposed framework with a publicly available traveling subject dataset from the Strategic Research Program for Brain Sciences (SRPBS) and harmonized the gray matter volume maps derived from eight source sites to a target site. The proposed framework significantly eliminated intersite differences in gray matter volumes. The embedded encoders successfully captured both the abstract textures of site factors and the concrete brain features. Moreover, the proposed framework exhibited outstanding performance relative to conventional statistical harmonization methods in terms of site effect removal, data distribution homogenization, and intrasubject similarity improvement. Finally, the proposed harmonization network provided fixable expandability, through which new sites could be linked to the target site via indirect schema without retraining the whole model. Together, the proposed method offers a powerful and interpretable deep learning-based harmonization framework for multisite neuroimaging data that can enhance reliability and reproducibility in multisite studies regarding brain development and brain disorders.
Display omitted
Although several transcriptome-wide association studies (TWASs) have been performed to identify genes associated with overall breast cancer (BC) risk, only a few TWAS have explored the differences in ...estrogen receptor-positive (ER+) and estrogen receptor-negative (ER-) breast cancer. Additionally, these studies were based on gene expression prediction models trained primarily in breast tissue, and they did not account for alternative splicing of genes.
In this study, we utilized two approaches to perform multi-tissue TWASs of breast cancer by ER subtype: (1) an expression-based TWAS that combined TWAS signals for each gene across multiple tissues and (2) a splicing-based TWAS that combined TWAS signals of all excised introns for each gene across tissues. To perform this TWAS, we utilized summary statistics for ER + BC from the Breast Cancer Association Consortium (BCAC) and for ER- BC from a meta-analysis of BCAC and the Consortium of Investigators of Modifiers of BRCA1 and BRCA2 (CIMBA).
In total, we identified 230 genes in 86 loci that were associated with ER + BC and 66 genes in 29 loci that were associated with ER- BC at a Bonferroni threshold of significance. Of these genes, 2 genes associated with ER + BC at the 1q21.1 locus were located at least 1 Mb from published GWAS hits. For several well-studied tumor suppressor genes such as TP53 and CHEK2 which have historically been thought to impact BC risk through rare, penetrant mutations, we discovered that common variants, which modulate gene expression, may additionally contribute to ER + or ER- etiology.
Our study comprehensively examined how differences in common variation contribute to molecular differences between ER + and ER- BC and introduces a novel, splicing-based framework that can be used in future TWAS studies.
