Search

Aim. We sought to profile gut microbiota’s role in combination of Bu Fei Hua Yu (BFHY) with cisplatin treatment. Methods. Non-small cell lung cancer (NSCLC) mice model were constructed followed by treatment with cisplatin alone or combined with BFHY. Mice weight and tumor volume were measured during the experiment. And mice cecum were detected by hematoxylin and eosin, cecum contents were collected for Enzyme Linked ImmuneSorbent Assay, and stool were profiled for metagenomic sequencing. Results. Combination of BFHY with cisplatin treatment decreased the tumor growth and relieved the damage of cecum. Expressions of interleukin-6 (IL-6), interleukin-1β (IL-1β), monocyte chemotactic protein 1 (MCP), and interferon-γ (IFN-γ) were decreased compared with cisplatin treatment alone. Linear discriminant analysis effect size analysis showed that g_Parabacteroides was downregulated and g_Escherichia and g_Blautia were upregulated after cisplatin treatment. After combination with BFHY, g_Bacteroides and g_Helicobacter were decreased. g_Klebsiella, g_Unclssified_Proteobacteria, and g_Unclssified_Clostridiates were increased. Moreover, heatmap results showed that Bacteroides abundance was increased significantly after cisplatin treatment; BFHY combination treatment reversed this state. Function analysis revealed that multiple functions were slightly decreased in cisplatin treatment alone and increased significantly after combination with BFHY. Conclusion. Our study provided evidence of an efficacy of combination of BFHY with cisplatin on treatment of NSCLC and revealed that gut microbiota plays a role in it. The above results provide new ideas on NSCLC treatment.

Visual retrieval tasks such as image retrieval and person re-identification (Re-ID) aim at effectively and thoroughly searching images with similar content or the same identity. After obtaining retrieved examples, re-ranking is a widely adopted post-processing step to reorder and improve the initial retrieval results by making use of the contextual information from semantically neighboring samples. Prevailing re-ranking approaches update distance metrics and mostly rely on inefficient crosscheck set comparison operations while computing expanded neighbors based distances. In this work, we present an efficient re-ranking method which refines initial retrieval results by updating features. Specifically, we reformulate re-ranking based on Graph Convolution Networks (GCN) and propose a novel Graph Convolution based Re-ranking (GCR) for visual retrieval tasks via feature propagation. To accelerate computation for large-scale retrieval, a decentralized and synchronous feature propagation algorithm which supports parallel or distributed computing is introduced. In particular, the plain GCR is extended for cross-camera retrieval and an improved feature propagation formulation is presented to leverage affinity relationships across different cameras. It is also extended for video-based retrieval, and Graph Convolution based Re-ranking for Video (GCRV) is proposed by mathematically deriving a novel profile vector generation method for the tracklet. Without bells and whistles, the proposed approaches achieve state-of-the-art performances on seven benchmark datasets from three different tasks, i.e., image retrieval, person Re-ID and video-based person Re-ID., Code is publicly available: https://github.com/WesleyZhang1991/GCN_rerank

The miniaturization of silicon-based electronics has motivated considerable efforts in exploring new electronic materials, including two-dimensional semiconductors and halide perovskites, which are usually too delicate to maintain their intrinsic properties during the harsh device fabrication steps. Here we report a convenient plug-and-probe approach for one-step simultaneous van der Waals integration of high-k dielectrics and contacts to enable top-gated transistors with atomically clean and electronically sharp dielectric and contact interfaces. By applying the plug-and-probe top-gate transistor stacks on two-dimensional semiconductors, we demonstrate an ideal subthreshold swing of 60 mV per decade. Using this approach on delicate lead halide perovskite, we realize a high-k top-gate CsPbBr

Aspergillus nidulans has been more extensively characterized than other Aspergillus species considering its morphology, physiology, metabolic pathways, and genetic regulation. As it has a rapid growth rate accompanied by simple nutritional requirements and a high tolerance to extreme cultural conditions, A. nidulans is a promising microbial cell factory to biosynthesize various products in industry. However, it remains unclear for whether it is also a suitable host for synthesizing abundant L-malic acid. In this study, we developed a convenient and efficient double-gene-editing system in A. nidulans strain TN02A7 based on the CRISPR–Cas9 and Cre-loxP systems. Using this gene-editing system, we made a L-malic acid-producing strain, ZQ07, derived from TN02A7, by deleting or overexpressing five genes (encoding Pyc, pyruvate carboxylase; OahA, oxaloacetate acetylhydrolase; MdhC, malate dehydrogenase; DctA, C4-dicarboxylic acid transporter; and CexA, citric acid transporter). The L-malic acid yield in ZQ07 increased to approximately 9.6 times higher (up to 30.7 g/L titer) than that of the original unedited strain TN02A7, in which the production of L-malic acid was originally very low. The findings in this study not only demonstrate that A. nidulans could be used as a potential host for biosynthesizing organic acids, but also provide a highly efficient gene-editing strategy in filamentous fungi.

Tripterygium Glycosides Tablets (TGT) has shown obvious anti-rheumatoid arthritis (RA) effects accompanied by hepatotoxicity. Despite that many studies looked at TGT's anti-RA or hepatotoxic mechanism and substance basis, the results were still insufficient. Furthermore, the anti-RA and hepatotoxicity investigations of TGT were undertaken separately, neglecting the relationship between efficacy and toxicity. Herein, an integrated approach combining metabolomics, network pharmacology, serum pharmacochemistry, and molecular docking was adopted to elucidate the mechanism and substance basis of Tripterygium Glycosides Tablets (TGT) on anti-rheumatoid arthritis and hepatotoxicity simultaneously. The results showed that 33 components in TGT were absorbed into rat serum. Two toxic targets (PRKCA, FASN), three effective targets (PLA2G10, PTGES, PLA2G1B), and four effective and toxic targets (PTGS1, PTGS2, PLA2G2A, ALOX5) were obtained by metabolomics combined with network analysis and network pharmacology. A component-target-RA-hepatotoxicity network was constructed and five hepatotoxic components (1-desacetylwilforgine, wilfordconine, wilforgine, wilformine, wilfornine D), eight effective-toxic components (14-oxo-19-(4 → 3) abeo-abieta-3,8,12-tetraen-19,18-olide, 7-oxo-18(4 → 3) abeo-abieta-3,8,11,13-tetraen-18-oic acid, hypoglaulide, triptotriterpenic acid A, wilforol F, wilforlide B, triptoquinone B, wilforlide A); and 23 non-effective and non-toxic components were acquired and validated by molecular docking. In addition, our research revealed that glycerophospholipid metabolism and ether lipid metabolism were correlated to both hepatotoxicity and anti-RA of TGT. While in sphingolipid metabolism, ceramidases regulated ceramide-sphingosine and phytoceramide-phytosphingosine reaction were found to be correlated to hepatotoxicity, sphinganine-1-phosphate lyase (SPL) regulated sphingosine 1-phosphate (S1P)-phosphoethanolamine and sphinganine 1-phosphate-phosphoethanolamine were found to be attributed to anti-RA effects.

Large language models (LLMs) have demonstrated impressive zero-shot abilities on a variety of open-ended tasks, while recent research has also explored the use of LLMs for multi-modal generation. In this study, we introduce mPLUG-Owl, a novel training paradigm that equips LLMs with multi-modal abilities through modularized learning of foundation LLM, a visual knowledge module, and a visual abstractor module. This approach can support multiple modalities and facilitate diverse unimodal and multimodal abilities through modality collaboration. The training paradigm of mPLUG-Owl involves a two-stage method for aligning image and text, which learns visual knowledge with the assistance of LLM while maintaining and even improving the generation abilities of LLM. In the first stage, the visual knowledge module and abstractor module are trained with a frozen LLM module to align the image and text. In the second stage, language-only and multi-modal supervised datasets are used to jointly fine-tune a low-rank adaption (LoRA) module on LLM and the abstractor module by freezing the visual knowledge module. We carefully build a visually-related instruction evaluation set OwlEval. Experimental results show that our model outperforms existing multi-modal models, demonstrating mPLUG-Owl's impressive instruction and visual understanding ability, multi-turn conversation ability, and knowledge reasoning ability. Besides, we observe some unexpected and exciting abilities such as multi-image correlation and scene text understanding, which makes it possible to leverage it for harder real scenarios, such as vision-only document comprehension. Our code, pre-trained model, instruction-tuned models, and evaluation set are available at https://github.com/X-PLUG/mPLUG-Owl. The online demo is available at https://www.modelscope.cn/studios/damo/mPLUG-Owl., Working in Process