Identification of a glioma functional network from gene fitness data using machine learning.

Glioblastoma multiforme (GBM) is an aggressive form of brain tumours that remains incurable despite recent advances in clinical treatments. Previous studies have focused on sub-categorizing patient samples based on clustering various transcriptomic data. While functional genomics data are rapidly accumulating, there exist opportunities to leverage these data to decipher glioma-associated biomarkers. We sought to implement a systematic approach to integrating data from high throughput CRISPR-Cas9 screening studies with machine learning algorithms to infer a glioma functional network. We demonstrated the network significantly enriched various biological pathways and may play roles in glioma tumorigenesis. From densely connected glioma functional modules, we further predicted 12 potential Wnt/ß-catenin signalling pathway targeted genes, including AARSD1, HOXB5, ITGA6, LRRC71, MED19, MED24, METTL11B, SMARCB1, SMARCE1, TAF6L, TENT5A and ZNF281. Cox regression modelling with these targets was significantly associated with glioma overall survival prognosis. Additionally, TRIB2 was identified as a glioma neoplastic cell marker in single-cell RNA-seq of GBM samples. This work establishes novel strategies for constructing functional networks to identify glioma biomarkers for the development of diagnosis and treatment in clinical practice.

Ecological function-oriented vegetation protection and restoration strategies in China's Loess Plateau.

Potential natural vegetation (PNV) can provide a reference for vegetation protection and restoration. Previous studies often used PNV patterns as a reference; however, they ignored PNV ecological functions, impeding the establishment of function-oriented vegetation protection and restoration plans. To address this issue, this study used Loess Plateau of China as a case study to propose an ecological function-oriented vegetation protection and restoration framework based on PNV patterns and ecological functions. The results showed that PNV patterns, ecological functions, and their synergistic and trade-off relationships represented distinct spatial differences that would be largely influenced by climate change. This suggested that vegetation protection and restoration should be adapted to climate change. The protection and potential restoration regions for actual forest and grass were detected based on the stable PNV regions. Approximately 34.5%-41.4% of actual forest and 81.2%-82.3% of actual grass should be protected. Further, 13.9%-16.2% of actual forest and 14.7%-15.2% of actual grass have the potential to be restored to grass and forest, respectively, and lastly, the priority regions of forest and grass protection and potential restoration were determined according to a composite ecological functions index. Moreover, forest protection should be prioritized, followed by forest potential restoration, grass potential restoration, and grass protection. These results would be conducive to forest and grass protection and restoration of the Loess Plateau. The proposed framework is applicable to other regions of the world for developing vegetation protection and restoration strategies.

Token Selection is a Simple Booster for Vision Transformers.

Vision transformers have recently attained state-of-the-art results in visual recognition tasks. Their success is largely attributed to the self-attention component, which models the global dependencies among the image patches (tokens) and aggregates them into higher-level features. However, self-attention brings significant training difficulties to ViTs. Many recent works thus develop various new self-attention components to alleviate this issue. In this article, instead of developing complicated self-attention mechanism, we aim to explore simple approaches to fully release the potential of the vanilla self-attention. We first study the token selection behavior of self-attention and find that it suffers from a low diversity due to attention over-smoothing, which severely limits its effectiveness in learning discriminative token features. We then develop simple approaches to enhance selectivity and diversity for self-attention in token selection. The resulted token selector module can server as a drop-in module for various ViT backbones and consistently boost their performance. Significantly, they enable ViTs to achieve 84.6% top-1 classification accuracy on ImageNet with only 25M parameters. When scaled up to 81M parameters, the result can be further improved to 86.1%. In addition, we also present comprehensive experiments to demonstrate the token selector can be applied to a variety of transformer-based models to boost their performance for image classification, semantic segmentation and NLP tasks. Code is available at https://github.com/zhoudaquan/dvit_repo.

Factorization Vision Transformer: Modeling Long-Range Dependency With Local Window Cost.

Transformers have astounding representational power but typically consume considerable computation which is quadratic with image resolution. The prevailing Swin transformer reduces computational costs through a local window strategy. However, this strategy inevitably causes two drawbacks: 1) the local window-based self-attention (WSA) hinders global dependency modeling capability and 2) recent studies point out that local windows impair robustness. To overcome these challenges, we pursue a preferable trade-off between computational cost and performance. Accordingly, we propose a novel factorization self-attention (FaSA) mechanism that enjoys both the advantages of local window cost and long-range dependency modeling capability. By factorizing the conventional attention matrix into sparse subattention matrices, FaSA captures long-range dependencies, while aggregating mixed-grained information at a computational cost equivalent to the local WSA. Leveraging FaSA, we present the factorization vision transformer (FaViT) with a hierarchical structure. FaViT achieves high performance and robustness, with linear computational complexity concerning input image spatial resolution. Extensive experiments have shown FaViT's advanced performance in classification and downstream tasks. Furthermore, it also exhibits strong model robustness to corrupted and biased data and hence demonstrates benefits in favor of practical applications. In comparison to the baseline model Swin-T, our FaViT-B2 significantly improves classification accuracy by 1% and robustness by 7% , while reducing model parameters by 14% . Our code will soon be publicly available: at https://github.com/q2479036243/FaViT.

Progressive Tandem Learning for Pattern Recognition With Deep Spiking Neural Networks.

Spiking neural networks (SNNs) have shown clear advantages over traditional artificial neural networks (ANNs) for low latency and high computational efficiency, due to their event-driven nature and sparse communication. However, the training of deep SNNs is not straightforward. In this paper, we propose a novel ANN-to-SNN conversion and layer-wise learning framework for rapid and efficient pattern recognition, which is referred to as progressive tandem learning. By studying the equivalence between ANNs and SNNs in the discrete representation space, a primitive network conversion method is introduced that takes full advantage of spike count to approximate the activation value of ANN neurons. To compensate for the approximation errors arising from the primitive network conversion, we further introduce a layer-wise learning method with an adaptive training scheduler to fine-tune the network weights. The progressive tandem learning framework also allows hardware constraints, such as limited weight precision and fan-in connections, to be progressively imposed during training. The SNNs thus trained have demonstrated remarkable classification and regression capabilities on large-scale object recognition, image reconstruction, and speech separation tasks, while requiring at least an order of magnitude reduced inference time and synaptic operations than other state-of-the-art SNN implementations. It, therefore, opens up a myriad of opportunities for pervasive mobile and embedded devices with a limited power budget.

A Multifunctional AIE Nanoprobe as a Drug Delivery Bioimaging and Cancer Treatment System.

Of all malignant brain tumors, glioma is the deadliest and most common, with a poor prognosis. Drug therapy is considered as a promising way to stop the progression of disease and even cure tumors. However, the presence of blood brain barrier (BBB) and blood tumor barrier (BTB) limits the delivery of these therapeutic genes. In this work, an intelligent cell imaging and cancer therapy drug delivery system targeting the blood-brain barrier and the highly expressed transferrin receptors (TfR) in gliomas has been successfully constructed, and an amphiphilic polymer (PLA-PEG-T7/TPE) with aggregation-induced emission (AIE) properties has been designed and successfully synthesized. PLA-PEG-T7/TPE self-assembled polymer micelles showed significant AIE effect in aqueous solution with good biocompatibility. Therefore, it can be used for potential biological imaging applications. In addition, drug-carrying micelles showed typical behavior of regulating drug release. Inhibition of cell proliferation in vitro showed that the drug-loaded micelles had dose-dependent cytotoxicity to LN229 cells. In the in vivo anti-tumor experiment, PLA-PEG-T7/TPE/TMZ had the best therapeutic effect. These results indicated that T7 functionalized PLA-PEG was a promising platform for nasopharyngeal cancer drug combination therapy.

Correlation of MMP-9 gene polymorphisms with aneurysmal subarachnoid hemorrhage and its prognosis.

Background: Aneurysmal subarachnoid hemorrhage (aSAH)-associated gene polymorphism is of great significance for the accurate diagnosis and individualized treatment of aSAH. This study aims to investigate the expression of matrix metalloproteinase-9 (MMP-9) gene in the peripheral blood of patients with aneurysmal subarachnoid hemorrhage (aSAH) and explore the correlations of MMP-9 polymorphisms with the onset and prognosis of the disease. Methods: A total of 80 aSAH patients (aSAH group) and 24 healthy (control group) people receiving physical examination were enrolled in the study. Western blotting was applied to detect the expression of MMP-9 gene in the peripheral blood in aSAH patients and healthy people. The genotyping of single nucleotide polymorphisms (rs42512, rs56212 and rs61221) in the promoter region of MMP-9 gene was analyzed by means of conformation-difference gel electrophoresis. Chi-square test was applied to examine the applicability of the distribution frequency of MMP-9 genotypes with genetic equilibrium law. The correlations of MMP-9 alleles and gene polymorphisms with the onset and prognosis of aSAH were determined. Results: The expression of MMP-9 protein in aSAH group was significantly higher than that in control group (P<0.05). The Hardy-Weinberg equilibrium analysis showed that MMP-9 gene polymorphisms were in agreement with the genetic equilibrium law. According to the results of genetic association analysis, only the polymorphism rs42512 and its alleles were significantly correlated with the onset and prognosis of aSAH (P<0.05). However, polymorphisms rs56212 and rs61221 and their alleles had no association with the onset and prognosis of aSAH (P>0.05). Conclusion: The polymorphism rs42512 in the promoter region of MMP-9 gene is related to the onset of aSAH, which provides further evidence for the diagnosis of aSAH.

A GLO10 score for the prediction of prognosis in high grade gliomas.

Gliomas are the most common lethal brain tumours and remain great heterogeneity in terms of histopathology and clinical outcomes. Among them, glioblastomas are the most aggressive tumours that lead to a median of less than one-year survival in patients. Despite the little improvement of in diagnosis and treatments for last decades, there is an urgent need for prognostic markers to distinguish high- and low-risk patients before treatment.Here, we generated a list of genes associated with glioblastoma progressions and then performed a comprehensive statistical modelling strategy to derive a 10-gene (GLO10) score from genome wide expression profiles of a large glioblastoma cohort (n=844). Our study demonstrated that the GLO10 score could successfully distinguish high- and low-risk patients with glioblastomas regardless their traditional pathological factors. Validated in four independent cohorts, the utility of GLO10 score could provide clinicians a robust prognostic prediction tool to assess risk levels upfront treatments.

Neurocytoma of the pituitary gland: A case report and literature review.

Extraventricular neurocytomas (EVN) are rare central nervous system tumors, often located within the cerebral hemisphere. The present study reports a case of a 56-year-old male patient with bitemporal hemianopsia. Computed tomography and magnetic resonance imaging revealed a tumor in the sellar region. The tumor was totally excised. Postoperative histological examination of the tumor demonstrated that synaptophysin, chromogranin-A and neuron-specific enolase were positive, while luteinizing hormone, follicle-stimulating hormone, growth hormone, prolactin, adrenocorticotropic hormone, thyroid-stimulating hormone, glial fibrillary acidic protein, S-100, nestin and epithelial membrane antigen were negative, which were the main pathological features of neurocytomas. This is the fourth case of EVN located in the sellar region reported. The associated studies are also reviewed.

Gene expression profile for predicting survival of patients with meningioma.

Current staging methods are inadequate for predicting the overall survival of meningioma. DNA microarray technologies improve the understanding of tumour progression. We analysed genome wide expression profiles of 119 meningioma samples from two previous published DNA microarray studies. The Cox proportional hazards regression models were applied to identify overall survival related gene signature. A total of 449 genes (109 upregulated and 340 downregulated) were identified as differentially expressed in meningioma. Among these differentially expressed genes, 37 genes were identified to be related to meningioma overall survival. Our 37-gene signature is closely associated with overall survival among patients with meningioma. This gene expression profile could provide an optimization of the clinical management and development of new therapeutic strategies for meningioma.