First Report of Tomato Yellow Mottle-Associated Virus Infecting Tobacco in China.

Tomato yellow mottle-associated virus (TYMaV) belongs to the genus Cytorhabdovirus in the family Rhabdoviridae and has been reported to infect a variety of Solanaceae crops, such as Solanum lycopersicum, S. nigrum, Capsicum annuum and Nicotiana benthamiana (Li et al. 2022, Li et al. 2023, Xu et al. 2017, Zhou et al. 2019). In August 2022, about 500 out of 2000 tobacco (N. tabacum) plants showing leaf distortion, crinkling and mosaic symptoms were found in one tobacco growing field in Xingren City, Guizhou Province, China. To identify the causal pathogen(s), leaves from 20 symptomatic tobacco plants were collected and pooled to perform small RNA deep sequencing (sRNA-Seq) and assembly. Briefly, total RNA was extracted with TRIzol Reagent (Takara, Kusatsu, Japan). A small RNA cDNA library was constructed by the small RNA Sample Pre Kit. sRNA-Seq was performed with an Illumina NovaSeq 6000 platform. About 29 million reads were obtained and 334 contigs generated after removal of host-derived sequences. Among them, 31 unique contigs mapped to the TYMaV genome (NC_034240.1), covering 28.43% of the genome with the mean read coverage of 0.92%. Meanwhile, 226 contigs mapped to the genome of a potyvirus, chilli veinal mottle virus (ChiVMV, NC_005778.1), covering 88.79% of the genome with the mean read coverage of 0.83%. To verify the sRNA-Seq result for TYMaV identification, reverse transcription (RT)- PCR was performed with specific primers TYMaV-F (5'-CTGACGTAGTGTTGGCAGAT-3') and TYMaV-R (5'-AACCTCCATGCAGAACCATGG-3'). The expected-size 936-bp fragment was amplified from total RNA of all 20 samples. Dot enzyme-linked immunosorbent assays (Dot-ELISA) with antibody for TYMaV (kindly provided by Dr. Zhenggang Li from Guangdong Academy of Agricultural Sciences) were performed and further verified TYMaV infection. In addition, five asymptomatic tobacco plants from the same field as controls were used to detect TYMaV by RT-PCR and Dot-ELISA, and all samples showed negative test results. Subsequently, 17 primer pairs (Supplementary Table 1) were used to obtain the full-length sequence of TYMaV from a single positive tobacco sample by RT-PCR, followed by Sanger sequencing at Sangon Biotech (Shanghai, China). The resulting amplicon sequences were assembled into a nearly full-length genome sequence of a TYMaV isolate from tobacco in Guizhou (TYMaV-GZ). BLASTn analysis of the 13, 393 nt-long sequence (GeneBank accession number, PP444718) revealed 84.7% and 87.2% nt sequence identity with the TYMaV tomato isolate (KY075646.1) and the TYMaV S. nigrum isolate (MW527091.1), respectively. Moreover, five S. nigrum plants showing leaf crinkling and mosaic symptoms from tobacco fields tested positive for TYMaV by RT-PCR assay, suggesting a potential spread of TYMaV between tobacco and S. nigrum, which may serve as a reservoir for the virus in the tobacco fields. However, the transmission route of TYMaV remains unknown, and further verification is needed. To our knowledge, this is the first report of TYMaV infecting tobacco crop in China. It will be important to assess the potential economic importance of TYMaV to tobacco production in China and elsewhere, and to elucidate the respective roles of this virus and ChiVMV in the leaf distorting and yellowing symptoms.

A tumor-targetable probe based on europium(III)/gadolinium(III) complex-conjugated transferrin for dual-modal time-gated luminescence and magnetic resonance imaging of cancerous cells in vitro and in vivo.

The synergy of magnetic resonance imaging (MRI) and time-gated luminescence imaging (TGLI) provides a robust platform with extensive spatial resolution (from submicrometer to hundred-micron) and unlimited penetration depth for visual detection of lesion tissues and target biomolecules. In this work, highly stable lanthanide (Eu3+ and Gd3+) complexes with a terpyridine polyacid ligand, CNSTTA-Ln3+, were chosen as signal reporters for TGLI (Ln3+ = Eu3+) and MRI (Ln3+ = Gd3+), respectively. After conjugating CNSTTA-Ln3+ with a tumor-targetable glycoprotein, transferrin (Tf), the obtained bioconjugate, showed low cytotoxicity and high stability and exhibited strong long-lived luminescence (Tf-CNSTTA-Eu3+, ϕ = 10.8%, τ = 1.27 ms), high magnetic resonance relaxivity (Tf-CNSTTA-Gd3+, r1 = 8.70 mM-1 s-1, r2 = 10.90 mM-1 s-1), and high binding affinity toward Tf receptor-overexpressed cancerous cells. On the basis of these features, a tumor-targetable probe was constructed by simply mixing Tf-CNSTTA-Eu3+ and Tf-CNSTTA-Gd3+, and successfully used for the bimodal TGLI and MRI of tumor cells in tumor-bearing mice. The bimodal imaging simultaneously provided the anatomical and molecular information of the tumor, which enabled the accuracy for tumor diagnosis to be mutually verified, and revealed the potential of Tf-CNSTTA-Gd3+/Eu3+ for the monitoring of cancer cells in vivo.

An activatable nanoprobe based on nanocomposites of visible-light-excitable europium(III) complex-anchored MnO2 nanosheets for bimodal time-gated luminescence and magnetic resonance imaging of tumor cells.

Bimodal imaging probes that combine magnetic resonance imaging (MRI) and photoluminescence imaging are quite appealing since they can supply both anatomical and molecular information to effectively ameliorate the accuracy of detection. In this study, an activatable nanoprobe, [Eu(BTD)3(DPBT)]@MnO2, for bimodal time-gated luminescence imaging (TGLI) and MRI has been constructed by anchoring visible-light-excitable Eu3+ complexes on lamellar MnO2 nanosheets. Due to the luminescence quenching effect and non-magnetic resonance (MR) activity of MnO2 nanosheets, the developed nanoprobe presents quite weak TGL and MR signals. After exposure to H2O2 or GSH, accompanied by the transformation from MnO2 to Mn2+, the nanoprobe exhibits rapid, sensitive, and selective "turn-on" responses towards GSH and H2O2 in TGL and MR detection modes. Furthermore, the nanoprobe displays high stability, low cytotoxicity, good biocompatibility and water dispersion. Given the high contents of GSH and H2O2 in cancer cells, the nanoprobe was used for the identification of cancer cells by TGLI of intracellular GSH and H2O2, as well as for the tracing of tumor cells in tumor-bearing mice by tumor-targeting in vivo MRI and TGLI of tumor tissues. The research outcomes proved the potential of [Eu(BTD)3(DPBT)]@MnO2 as a useful nanoprobe for the tracing and accurate detection of cancer cells in vitro and in vivo via bimodal TGLI and MRI.

Synergism between hierarchical MFI zeolites and alumina in alkene cross-metathesis reactions as a function of composition.

Synergism between hierarchical zeolites and alumina in the preparation of active Mo catalysts, as a function of composition ratios, has been demonstrated in the cross-metathesis reaction between ethene and 2-butene. The metathesis reaction activity, reflected by ethene conversion, increases from 24.1% to 49.2% with the increase in the alumina content in composites from 10 wt% to 30 wt%. A further increase in the alumina content leads to the reduction in the metathesis activity, in which the ethene conversion decreases from 30.3% to 4.8% upon the enhanced alumina content from 50 wt% to 90 wt%. The impact of alumina content on the metathesis activity is closely associated with the interaction mode between the hierarchical ZSM-5 zeolite and alumina. TEM observation as well as EDS analysis and XPS results prove the progressive coating of alumina phase on the surface of zeolites along with the progressive enhancement of alumina content. The moderate alumina content in the composite enables the desired interaction between hierarchical zeolites and alumina, which is beneficial for the preparation of active catalysts for the alkene cross-metathesis reaction.

Insight into the Evolution Track for the Metathesis of Alkenes within Hierarchical Zeolite-Based Catalysts.

The usage of hierarchical MFI zeolite enables a boost of the catalytic performance of Mo-based catalysts for the olefin-metathesis reaction. The harvest of active catalysts roots in a segmental evolution track between hierarchical zeolite and Al2 O3 slices for the fabrication of active sites. The working evolution track requires the indispensable engagements from intracrystalline mesoporous surface, Al2 O3 slices, and zeolitic Brønsted acid sites. The infilling of disaggregated Al2 O3 slices into the intracrystalline mesopores triggers the creation of localized intrazeolite-Al2 O3 interfaces, which enables the subsequent migration and trapping of surface molybdates into the micropores. The insulation of intrazeolite-Al2 O3 interface or shielding of zeolitic Brønsted acid sites leads to the break of the evolution track. Our findings disclose the hidden functionality of mesoporosity as intrazeolite interface boundary for the fabrication of active sites and supply a new strategy for the rational design of zeolite catalysts.

QBox: Partial Transfer Learning With Active Querying for Object Detection.

Object detection requires plentiful data annotated with bounding boxes for model training. However, in many applications, it is difficult or even impossible to acquire a large set of labeled examples for the target task due to the privacy concern or lack of reliable annotators. On the other hand, due to the high-quality image search engines, such as Flickr and Google, it is relatively easy to obtain resource-rich unlabeled datasets, whose categories are a superset of those of target data. In this article, to improve the target model with cost-effective supervision from source data, we propose a partial transfer learning approach QBox to actively query labels for bounding boxes of source images. Specifically, we design two criteria, i.e., informativeness and transferability, to measure the potential utility of a bounding box for improving the target model. Based on these criteria, QBox actively queries the labels of the most useful boxes from the source domain and, thus, requires fewer training examples to save the labeling cost. Furthermore, the proposed query strategy allows annotators to simply labeling a specific region, instead of the whole image, and, thus, significantly reduces the labeling difficulty. Extensive experiments are performed on various partial transfer benchmarks and a real COVID-19 detection task. The results validate that QBox improves the detection accuracy with lower labeling cost compared to state-of-the-art query strategies for object detection.

CCMN: A General Framework for Learning With Class-Conditional Multi-Label Noise.

Class-conditional noise commonly exists in machine learning tasks, where the class label is corrupted with a probability depending on its ground-truth. Many research efforts have been made to improve the model robustness against the class-conditional noise. However, they typically focus on the single label case by assuming that only one label is corrupted. In real applications, an instance is usually associated with multiple labels, which could be corrupted simultaneously with their respective conditional probabilities. In this paper, we formalize this problem as a general framework of learning with Class-Conditional Multi-label Noise (CCMN for short). We establish two unbiased estimators with error bounds for solving the CCMN problems, and further prove that they are consistent with commonly used multi-label loss functions. Finally, a new method for partial multi-label learning is implemented with the unbiased estimator under the CCMN framework. Empirical studies on multiple datasets and various evaluation metrics validate the effectiveness of the proposed method.

Two-Stage Stochastic Optimization for the Pre-Position and Reconfiguration of Microgrid Defense Resources against Natural Disasters.

With the aggravation and evolution of global warming, natural disasters such as hurricanes occur more frequently, posing a great challenge to large-scale power systems. Therefore, the pre-position and reconfiguration of the microgrid defense resources by means of Mobile Energy Storage Vehicles (MEVs) and tie lines in damaged scenarios have attracted more and more attention. This paper proposes a novel two-stage optimization model with the consideration of MEVs and tie lines to minimize the shed loads and the outage duration of loads according to their proportional priorities. In the first stage, tie lines addition and MEVs pre-position are decided prior to a natural disaster; in the second stage, the switches of tie lines and original lines are operated and MEVs are allocated from staging locations to allocation nodes according to the specific damaged scenarios after the natural disaster strikes. The proposed load restoration method exploits the benefits of MEVs and ties lines by microgrid formation to pick up more critical loads. The progressive hedging algorithm is employed to solve the proposed scenario-based two-stage stochastic optimization problem. Finally, the effectiveness and superiority of the proposed model and applied algorithm are validated on an IEEE 33-bus test case.

Realizing Fast Synthesis of High-Silica Zeolite Y with Remarkable Catalytic Performance.

High-silica zeolite Y (FAU) plays a vital role in (petro)chemical industries. However, the slow nucleation and growth kinetics of the high-silica FAU framework limit its direct synthesis and the improvement of framework SiO2 /Al2 O3 ratio (SAR). Here, a facile strategy is developed to realize the fast crystallization of high-silica zeolite Y, which involves the combination of high crystallization temperature, ultra-stable Y (USY) seeds and efficient organic-structure directing agent (OSDA). The synthesis can be finished in 5-16 h at 160 °C and with tunable SAR up to 18.2, and the key factors affecting crystallization kinetics and phase purity are elucidated. Moreover, the crystallization process was monitored to reveal the fast crystal growth mechanism. The high-silica products possess high (hydro)thermal stability and abundant strong acid sites, which endow them excellent catalytic cracking performance, obviously superior to commercial USY.

Partial Multi-Label Learning With Noisy Label Identification.

Partial multi-label learning (PML) deals with problems where each instance is assigned with a candidate label set, which contains multiple relevant labels and some noisy labels. Recent studies usually solve PML problems with the disambiguation strategy, which recovers ground-truth labels from the candidate label set by simply assuming that the noisy labels are generated randomly. In real applications, however, noisy labels are usually caused by some ambiguous contents of the example. Based on this observation, we propose a partial multi-label learning approach to simultaneously recover the ground-truth information and identify the noisy labels. The two objectives are formalized in a unified framework with trace norm and l1 norm regularizers. Under the supervision of the observed noise-corrupted label matrix, the multi-label classifier and noisy label identifier are jointly optimized by incorporating the label correlation exploitation and feature-induced noise model. Furthermore, by mapping each bag to a feature vector, we extend PML-NI method into multi-instance multi-label learning by identifying noisy labels based on ambiguous instances. A theoretical analysis of generalization bound and extensive experiments on multiple data sets from various real-world tasks demonstrate the effectiveness of the proposed approach.

Cascade MIR Ho:YLF laser at 2.1†µm and 2.9 µm.

Cascade transitions of Ho3+:5I6→5I7 and 5I7→5I8 provide a platform for a dual-wavelength mid-infrared (MIR) laser. In this paper, a continuous wave cascade MIR Ho:YLF laser operating at 2.1 and 2.9 µm is realized at room temperature. The total output power of 929 mW with 778 mW at 2.9 µm and 151 mW at 2.1 µm is obtained under the absorbed pump power of 5 W. Compared to the non-cascade mode, 2.9-µm lasing threshold is reduced by 10.3% and the slope efficiency is increased by 76.1% with the supports of cascade lasing at 2.1 µm. However, 2.9-µm lasing is the key population accumulation of the 5I7 level, which thus efficiently reduces the threshold and improves the output power of the 2.1-µm laser. Our results put forward a way for generating cascade dual-wavelength MIR lasing in Ho3+-doped crystals.

HIPPO signaling-related signature for predicting prognosis and therapeutic response in gastric cancer.

Background: Gastric cancer (GC) is a multifactorial progressive disease with high mortality and heterogeneous prognosis. Effective prognostic biomarkers for GC were critically needed. Hippo signaling pathway is one of the critical mechanisms regulating the occurrence and development of GC, and has potential clinical application value for the prognosis and treatment of GC patients. However, there is no effective signature based on Hippo signaling pathway-related genes (HSPRGs) to predict the prognosis and treatment response of GC patients. Our study aimed to build a HSPRGs signature and explore its performance in improving prognostic assessment and drug therapeutic response in GC. Methods: Based on gene expression profiles obtained from The Cancer Genome Atlas (TCGA) database, we identified differentially expressed HSPRGs and conducted univariate and the least absolute shrinkage and selection operator (LASSO) Cox regression analysis to construct a multigene risk signature. Subsequently, the Kaplan-Meier curve and receiver operating characteristic (ROC) were performed to evaluate the predictive value of the risk signature in both training and validation cohort. Furthermore, we carried out univariate and multivariate Cox regression analysis to investigate the independent prognostic factors and establish a predictive nomogram. The enriched signaling pathways in risk signature were analyzed by gene set enrichment analysis (GSEA). Tumor immune dysfunction and exclusion (TIDE) and drug sensitivity analysis were performed to depict therapeutic response in GC. Results: In total, 38 differentially expressed HSPRGs were identified, and final four genes (DLG3, TGFB3, TGFBR1, FZD6) were incorporated to build the signature. The ROC curve with average 1-, 3-, and 5-year areas under the curve (AUC) equal to .609, .634, and .639. Clinical ROC curve revealed that risk signature was superior to other clinicopathological factors in predicting prognosis. Calibration curves and C-index (.655) of nomogram showed excellent consistency. Besides, in the immunotherapy analysis, exclusion (p < 2.22 × 10-16) and microsatellite instability (p = .0058) performed significantly differences. Finally, our results suggested that patients in the high-risk group were more sensitive to specific chemotherapeutic agents. Conclusion: Results support the hypothesis that Hippo-related signature is a novel prognostic biomarker and predictor, which could help optimize GC prognostic stratification and inform clinical medication decisions.

Visual-guided attentive attributes embedding for zero-shot learning.

Zero-shot learning (ZSL) aims to learn a classifier for unseen classes by exploiting both training data from seen classes and external knowledge. In many visual tasks such as image classification, a set of high-level attributes that describe the semantic properties of classes are used as the external knowledge to bridge seen and unseen classes. While the attributes are usually treated equally by previous ZSL studies, we observe that the contribution of different attributes varies significantly over model training. To adaptively exploit the discriminative information embedded in different attributes, we propose a novel encoder-decoder framework with attention mechanism on the attribute level for zero-shot learning. Specifically, by mapping the visual features into a semantic space, the more discriminative attributes are emphasized with larger attention weights. Further, the attentive attributes and the class prototypes are simultaneously decoded to the visual space so that the hubness problem can be eased. Finally, the labels are predicted in the visual space. Extensive experiments on multiple benchmark datasets demonstrate that our proposed model achieves a significant boost over several state-of-the-art methods for ZSL task and comparative results for generalized ZSL task.

Nonlinear Optical Response of Gold Nanobipyramids for a Doubly Q-Switched Ho-Doped Laser at a Wavelength of 2.1 µm.

Gold nanobipyramids (Au-NBPs) were successfully fabricated using the seed-mediated growth method. The saturable absorption performance of the Au-NBPs at a 2-µm band wavelength was characterized. Using excellent-quality, mature Ho:YLF crystals, a doubly Q-switched (DQS) laser joining an acousto-optic modulator (AOM) with an Au-NBP saturable absorber (SA) was achieved. When the modulation rate of the AOM was 1 kHz, the shortest pulse width (54 ns) was attained, corresponding to the highest peak power (3.87 kW). This was compared with a singly Q-switched laser joining an AOM with an Au-NBP SA, whereby the maximum pulse width compression ratio was 15.2 and the highest peak power enhancement factor was 541.3. Our study has shown that Au-NBPs are a potential saturable absorption nanomaterial, and the DQS laser has the benefit of compressing the pulse width and increasing the peak power at a wavelength of 2.1 µm.

Recent Advances in Open Set Recognition: A Survey.

In real-world recognition/classification tasks, limited by various objective factors, it is usually difficult to collect training samples to exhaust all classes when training a recognizer or classifier. A more realistic scenario is open set recognition (OSR), where incomplete knowledge of the world exists at training time, and unknown classes can be submitted to an algorithm during testing, requiring the classifiers to not only accurately classify the seen classes, but also effectively deal with unseen ones. This paper provides a comprehensive survey of existing open set recognition techniques covering various aspects ranging from related definitions, representations of models, datasets, evaluation criteria, and algorithm comparisons. Furthermore, we briefly analyze the relationships between OSR and its related tasks including zero-shot, one-shot (few-shot) recognition/learning techniques, classification with reject option, and so forth. Additionally, we also review the open world recognition which can be seen as a natural extension of OSR. Importantly, we highlight the limitations of existing approaches and point out some promising subsequent research directions in this field.

A Bottom-Up Strategy for the Synthesis of Highly Siliceous Faujasite-Type Zeolite.

High-silica zeolite Y is a desired catalytic material for oil refining and the petrochemical industry. However, its direct synthesis remains a symbolic challenge in the field of zeolite synthesis, with a limited improvement of the framework SiO2 /Al2 O3 ratio (SAR) from ≈5 to 9 over the past 60 years. Here, the synthesis of highly siliceous zeolite Y with tunable SAR up to 15.6 through a cooperative strategy is reported, which involves the use of FAU nuclei, a bulky organic structure-directing agent (OSDA), and a gel system with low alkalinity (named NOA-co strategy). A series of quaternary alkylammonium ions is discovered as effective OSDAs based on the NOA-co strategy, and the relevant crystallization mechanism is elucidated. Moreover, the high-silica products are demonstrated to have greatly improved (hydro)thermal stability, high concentration of strong acid sites, and uniform acid distribution, which lead to superior catalytic performance in the cracking of bulky hydrocarbons. It is anticipated that this synthetic strategy will benefit the synthesis and development of zeolitic catalysts in a wide range of reaction processes.

Insights into the Resorcinol-Formaldehyde Resin Coating Process Focusing on Surface Modification of Colloidal SiO2 Particles.

This article provides a systematic study on the resorcinol-formaldehyde (RF) resin coating via a sol-gel process with focus on surface modification of the core. With colloidal SiO2 particles as the model core material, diverse surface modification methods were investigated to verify the feasibility of subsequent RF coating process. It is confirmed that the RF coating is strongly influenced by surface charge of the SiO2 core, which can be adjusted by suitable surface modification. Both cationic surfactant and amino functional group can modify the silica surface with positive charge, and it is readily coated with the negatively charged RF resin. The primary amine surfactant with positive charge fails to induce the RF coating onto the SiO2 particles, which may be due to the relatively weak interaction between the surfactant and silica. In addition, the negatively charged sulfydryl functionalization also contributes to a successful coating probably through the gathering of cationic ions around the core. The RF coating process has opened a versatile avenue for the construction of yolk-shell carbon-encapsulated nanocomposites. Catalytic tests indicate that the catalytic performances of the synthesized nanocomposites depend strongly on the method of synthesis.

Querying Representative and Informative Super-Pixels for Filament Segmentation in Bioimages.

Segmenting bioimage based filaments is a critical step in a wide range of applications, including neuron reconstruction and blood vessel tracing. To achieve an acceptable segmentation performance, most of the existing methods need to annotate amounts of filamentary images in the training stage. Hence, these methods have to face the common challenge that the annotation cost is usually high. To address this problem, we propose an interactive segmentation method to actively select a few super-pixels for annotation, which can alleviate the burden of annotators. Specifically, we first apply a Simple Linear Iterative Clustering (i.e., SLIC) algorithm to segment filamentary images into compact and consistent super-pixels, and then propose a novel batch-mode based active learning method to select the most representative and informative (i.e., BMRI) super-pixels for pixel-level annotation. We then use a bagging strategy to extract several sets of pixels from the annotated super-pixels, and further use them to build different Laplacian Regularized Gaussian Mixture Models (Lap-GMM) for pixel-level segmentation. Finally, we perform the classifier ensemble by combining multiple Lap-GMM models based on a majority voting strategy. We evaluate our method on three public available filamentary image datasets. Experimental results show that, to achieve comparable performance with the existing methods, the proposed algorithm can save 40 percent annotation efforts for experts.

Fast Multi-Instance Multi-Label Learning.

In many real-world tasks, particularly those involving data objects with complicated semantics such as images and texts, one object can be represented by multiple instances and simultaneously be associated with multiple labels. Such tasks can be formulated as multi-instance multi-label learning (MIML) problems, and have been extensively studied during the past few years. Existing MIML approaches have been found useful in many applications; however, most of them can only handle moderate-sized data. To efficiently handle large data sets, in this paper we propose the MIMLfast approach, which first constructs a low-dimensional subspace shared by all labels, and then trains label specific linear models to optimize approximated ranking loss via stochastic gradient descent. Although the MIML problem is complicated, MIMLfast is able to achieve excellent performance by exploiting label relations with shared space and discovering sub-concepts for complicated labels. Experiments show that the performance of MIMLfast is highly competitive to state-of-the-art techniques, whereas its time cost is much less. Moreover, our approach is able to identify the most representative instance for each label, and thus providing a chance to understand the relation between input patterns and output label semantics.

Mesopore Engineering for Well-Defined Mesoporosity in Al-Rich Aluminosilicate Zeolites.

Desilication has been proven an effective approach for the construction of well-defined hierarchical porosities inside zeolites with an optimal framework Al content (Si/Al=25-50). However, for the Al-rich aluminosilicate zeolites, desilication is constrained by the excess and extensive shielding effects from high Al-contents. The developments in the desilication of siliceous zeolites convey a simplified principle of controlled dissolution of the microporous matrix for the construction of hierarchical porosities, which benefits the innovation of synthetic approaches for Al-rich zeolites. The perturbations to the environments of framework Al species may alleviate the excess shielding effects. This review highlights two corresponding protocols of sequential "fluorination-desilication" and "steaming-desilication" for the construction of hierarchical porosities inside Al-rich ZSM-5 zeolites. The success of these two protocols revitalizes the prevailing understanding of the interplay between dealumination and desilication, and implies the necessity of investigating the overlooked roles of extra-framework Al species. Despite the long history and significant achievements in the last decade, fundamental understandings at the molecule level are still limited for the desilication-based top-down approaches. In particular, the investigations on Al-rich zeolites just find their growing. The bridging of dealumination and desilication is essential for other industrially relevant Al-rich zeolites (e.g., faujasite zeolites). The complexities in the inherent characters (topology, spatial distribution, proximity, etc.) and apparent parameters (morphology, crystal/particle size, etc.) demand constructive synthetic toolboxes and further fundamental understanding.