Explore the vast range of titles available.

Circular RNAs (circRNAs) are a new class of non-coding RNAs formed by covalently closed loops through backsplicing. Recent methodologies have enabled in-depth characterization of circRNAs for identification and potential functions. CircRNAs play important roles in various biological functions as microRNA sponges, transcriptional regulators and combining with RNA binding proteins. Recent studies indicated that some cytoplasmic circRNAs can be effectively translated into detectable peptides, which enlightened us on the importance of circRNAs in cellular physiology function. Internal Ribosome Entry site (IRES)- and N 6 -methyladenosines (m 6 A)-mediated cap-independent translation initiation have been suggested to be potential mechanism for circRNA translation. To date, several translated circRNAs have been uncovered to play pivotal roles in human cancers. In this review, we introduced the properties and functions of circRNAs, and characterized the possible mechanism of translation initiation and complexity of the translation ability of circRNAs. We summarized the emerging functions of circRNA-encoded proteins in human cancer. The works on circRNA translation will open a hidden human proteome, and enhance us to understand the importance of circRNAs in human cancer, which has been poorly explored so far.

Background Hibernation is a survival mechanism in the winter for some animals. Fat preserved instead of glucose produced is the primary fuel during winter hibernation of mammals. Many genes involved in lipid metabolism are regulated by the peroxisome proliferator-activated receptor alpha (PPARα). The role of PPARα in hibernation of mammals remains largely unknown. Using a multidisciplinary approach, we investigated whether PPARα is adapted to hibernation in bats. Results Evolutionary analyses revealed that the ω value of Ppar α of the ancestral lineage of hibernating bats in both Yinpterochiroptera and Yangochiroptera was lower than that of non-hibernating bats in Yinpterochiroptera, suggesting that a higher selective pressure acts on Ppar α in hibernating bats. PPARα expression was found to be increased at both mRNA and protein levels in distantly related bats ( Rhinolophus ferrumequinum and Hipposideros armiger in Yinpterochiroptera and Myotis ricketti in Yangochiroptera) during their torpid episodes. Transcription factors such as FOXL1, NFYA, NFYB, SP1, TBP, and ERG were bioinformatically determined to have a higher binding affinity to the potential regulatory regions of Ppar α in hibernating than in non-hibernating mammals. Genome-wide bioinformatic analyses of 64 mammalian species showed that PPARα has more potential target genes and higher binding affinity to these genes in hibernating than in non-hibernating mammals. Conclusions We conclude that PPARα is adapted to hibernation in bats based on the observations that Ppar α has a more stringent functional constraint in the ancestral lineage of hibernating bats and a higher level of expression in hibernating than in non-hibernating bats. We also conclude that PPARα plays a very important role in hibernation as hibernators have more PPARα target genes than non-hibernators, and PPARα in hibernators has a higher binding affinity for its target genes than in non-hibernators.

In order to illustrate the epigenetic heterogeneity, versatile tools of single-cell ChIP-seq (scChIP-seq) are necessary to meet the convenience and accuracy. Here, we develop MobiChIP, a compatible ChIP-seq library construction method based current sequencing platform with single cell level. As a novel capture strategy, MobiChIP is efficient to capture the fragments from tagmented nuclei of numerous species and execute the mixing of samples from different tissues or species. Especially, this strategy enables the flexible sequencing manipulation and sufficient nucleosome amplification without customized sequencing primers. MobiChIP reveals the landscape of chromatin regulation regions with active (H3K27ac) and repressive (H3K27me3) histone modification markers in peripheral blood mononuclear cells (PBMCs), and accurately unveiled the epigenetic repression of hox gene cluster in PBMCs than ATAC-seq. Meanwhile, we complete the bioinformatics pipeline to integrates the scChIP-seq data and scRNA-seq to illustrate the cellular epigenetic and genetic heterogeneity.Competing Interest StatementThe authors have declared no competing interest.

Diffusion Transformer (DiT) models have achieved unprecedented quality in image and video generation, yet their iterative sampling process remains computationally prohibitive. To accelerate inference, feature caching methods have emerged by reusing intermediate representations across timesteps. However, existing caching approaches treat all feature components uniformly. We reveal that DiT feature spaces contain distinct principal and residual subspaces with divergent temporal behavior: the principal subspace evolves smoothly and predictably, while the residual subspace exhibits volatile, low-energy oscillations that resist accurate prediction. Building on this insight, we propose SVD-Cache, a subspace-aware caching framework that decomposes diffusion features via Singular Value Decomposition (SVD), applies exponential moving average (EMA) prediction to the dominant low-rank components, and directly reuses the residual subspace. Extensive experiments demonstrate that SVD-Cache achieves near-lossless across diverse models and methods, including 5.55\\(\\times\\) speedup on FLUX and HunyuanVideo, and compatibility with model acceleration techniques including distillation, quantization and sparse attention. Our code is in supplementary material and will be released on Github.

Diffusion Transformers achieve impressive generative quality but remain computationally expensive due to iterative sampling. Recently, dynamic resolution sampling has emerged as a promising acceleration technique by reducing the resolution of early sampling steps. However, existing methods rely on heuristic re-noising at every resolution transition, injecting noise that breaks cross-stage consistency and forces the model to relearn global structure. In addition, these methods indiscriminately upsample the entire latent space at once without checking which regions have actually converged, causing accumulated errors, and visible artifacts. Therefore, we propose \\textbf{Fresco}, a dynamic resolution framework that unifies re-noise and global structure across stages with progressive upsampling, preserving both the efficiency of low-resolution drafting and the fidelity of high-resolution refinement, with all stages aligned toward the same final target. Fresco achieves near-lossless acceleration across diverse domains and models, including 10\\(\\times\\) speedup on FLUX, and 5\\(\\times\\) on HunyuanVideo, while remaining orthogonal to distillation, quantization and feature caching, reaching 22\\(\\times\\) speedup when combined with distilled models. Our code is in supplementary material and will be released on Github.

Diffusion Transformers offer state-of-the-art fidelity in image and video synthesis, but their iterative sampling process remains a major bottleneck due to the high cost of transformer forward passes at each timestep. To mitigate this, feature caching has emerged as a training-free acceleration technique that reuses or forecasts hidden representations. However, existing methods often apply a uniform caching strategy across all feature dimensions, ignoring their heterogeneous dynamic behaviors. Therefore, we adopt a new perspective by modeling hidden feature evolution as a mixture of ODEs across dimensions, and introduce HyCa, a Hybrid ODE solver inspired caching framework that applies dimension-wise caching strategies. HyCa achieves near-lossless acceleration across diverse domains and models, including 5.55 times speedup on FLUX, 5.56 times speedup on HunyuanVideo, 6.24 times speedup on Qwen-Image and Qwen-Image-Edit without retraining.

目的探讨不同病理分期糖尿病肾病（DKD）患者肾小球中ROBO4/ARF6的表达情况及其与肾小球内皮细胞（MGECs）通透性的相关性。方法收集30例DKD患者肾穿刺活检标本,参考Tervaert病理分期将DKD肾组织分为早、中、晚期各10例,以10例正常肾组织标本作为对照。采用免疫组化法检测ROBO4/ARF6在MGECs中的表达,分析其与蛋白尿、血清肌酐、肾小球滤过率（e GFR）、糖化血红蛋白的相关性。结果正常MGECs中有丰富的ROBO4表达,而ARF6表达较少。DKD组ROBO4主要在MGECs中表达,而ARF6在MGECs和肾小管上皮细胞中均有表达。与正常对照组相比,DKD组ROBO4染色阳性强度明显降低（P〈0.05）,且随病理分期的增加阳性强度逐渐减弱;而ARF6阳性强度随DKD病理分期的增加逐渐增高（P〈0.05）。ROBO4在MGECs中的阳性强度与24h蛋白尿呈负相关（r=–0.840,P〈0.01）,与血清肌酐呈负相关（r=–0.689,P〈0.01）,与糖化血红蛋白呈负相关（r=–0.660,P〈0.01）,与e GFR呈正相关（r=0.589,P〈0.01）;ARF6在MGECs中的染色强度与血清肌酐、e GFR无相关性（P〉0.05）,与24h蛋白尿呈正相关（r=0.603,P〈0.01）,与糖化血红蛋白呈正相关（r=0.582,P〈0.01）。结论 DKD患者MGECs中存在ROBO4低表达和ARF6高表达,提示ROBO4/ARF6信号通路可能参与了DKD肾小球内皮、血管的病理损伤和尿蛋白的进展。