Diverse nucleotide substitutions in rice base editing mediated by novel TadA variants.

CRISPR-mediated base editors have been widely used to correct defective alleles and create novel alleles by artificial evolution in rapid genetic improvement of crops. The editing capabilities of base editors strictly rely on the performance of various nucleotide modification enzymes. Compared to the well-developed adenine base editors (ABEs), cytosine base editors (CBEs) and dual base editors suffer from unstable editing efficiency and pattern at different genomic loci in rice, significantly limiting their application. Here, the activities of multiple evolved TadA8e variants in base editing have been comprehensively examined in rice. We found that both TadA-CDd and TadA-E27R/N46L achieve more robust C-to-T editing than previously reported hyperactive hAID*Δ, whereas TadA-CDd outperformed TadA-E27R/N46L. Also, the C-to-G base editor (CGBE) engineered with TadA-CDd and OsUNG performs highly efficient C-to-G editing in rice compared to TadA-N46P. In addition, the dual base editor, which was constructed with a single protein TadDE, enables simultaneously edit C-to-T and A-to-G at high efficiency in rice. Collectively, our study demonstrates that TadA8e derivatives improve both CBE and dual base editor in rice, provide a powerful way in inducing diverse nucleotide substitutions for plant genome editing.

Rapid Sample Pretreatment Facilitating SERS Detection of Trace Weak Organic Acids/Bases in Complex Matrices.

Fast and efficient sample pretreatment is the prerequisite for realizing surface-enhanced Raman spectroscopy (SERS) detection of trace targets in complex matrices, which is still a big issue for the practical application of SERS. Recently, we have proposed a highly performed liquid-liquid extraction (LLE)-back extraction (BE) for weak acids/bases extraction in drinking water and beverage samples. However, the performance efficiency decreased drastically on facing matrices like food and biological blood. Based on the total interaction energies among target, interferent, and extractant molecules, solid-phase extraction (SPE) with a higher selectivity was introduced in advance of LLE-BE, which enabled the sensitive (µg L-1 level) and rapid (within 10 min) SERS detection of both koumine (a weak base) and celastrol (a weak acid) in different food and biological samples. Further, the high SERS sensitivity was determined unmanned by Vis-CAD (a machine learning algorithm), instead of the highly demanded expert recognition. The generality of SPE-LLE-BE for various weak acids/bases (2 < pKa < 12), accompanied by the high efficiency, easy operation, and low cost, offers SERS as a powerful on-site and efficient inspection tool in food safety and forensics.

Challenges and support needs in psychological and physical health among pilots: a qualitative study.

Introduction: Physical and mental health problems among pilots affect their working state and impact flight safety. Although pilots' physical and mental health problems have become increasingly prominent, their health has not been taken seriously. This study aimed to clarify challenges and support needs related to psychological and physical health among pilots to inform development of a more scientific and comprehensive physical and mental health system for civil aviation pilots. Methods: This qualitative study recruited pilots from nine civil aviation companies. Focus group interviews via an online conference platform were conducted in August 2022. Colaizzi analysis was used to derive themes from the data and explore pilots' experiences, challenges, and support needs. Results: The main sub-themes capturing pilots' psychological and physical health challenges were: (1) imbalance between family life and work; (2) pressure from assessment and physical examination eligibility requirements; (3) pressure from worries about being infected with COVID-19; (4) nutrition deficiency during working hours; (5) changes in eating habits because of the COVID-19 pandemic; (6) sleep deprivation; (7) occupational diseases; (8) lack of support from the company in coping with stress; (9) pilots' yearly examination standards; (10) support with sports equipment; (11) respecting planned rest time; and (12) isolation periods. Discussion: The interviewed pilots experienced major psychological pressure from various sources, and their physical health condition was concerning. We offer several suggestions that could be addressed to improve pilots' physical and mental health. However, more research is needed to compare standard health measures for pilots around the world in order to improve their physical and mental health and contribute to overall aviation safety.

A Survey on 3D Skeleton-Based Action Recognition Using Learning Method.

Three-dimensional skeleton-based action recognition (3D SAR) has gained important attention within the computer vision community, owing to the inherent advantages offered by skeleton data. As a result, a plethora of impressive works, including those based on conventional handcrafted features and learned feature extraction methods, have been conducted over the years. However, prior surveys on action recognition have primarily focused on video or red-green-blue (RGB) data-dominated approaches, with limited coverage of reviews related to skeleton data. Furthermore, despite the extensive application of deep learning methods in this field, there has been a notable absence of research that provides an introductory or comprehensive review from the perspective of deep learning architectures. To address these limitations, this survey first underscores the importance of action recognition and emphasizes the significance of 3-dimensional (3D) skeleton data as a valuable modality. Subsequently, we provide a comprehensive introduction to mainstream action recognition techniques based on 4 fundamental deep architectures, i.e., recurrent neural networks, convolutional neural networks, graph convolutional network, and Transformers. All methods with the corresponding architectures are then presented in a data-driven manner with detailed discussion. Finally, we offer insights into the current largest 3D skeleton dataset, NTU-RGB+D, and its new edition, NTU-RGB+D 120, along with an overview of several top-performing algorithms on these datasets. To the best of our knowledge, this research represents the first comprehensive discussion of deep learning-based action recognition using 3D skeleton data.

Development and validation of a novel nomogram for predicting good neoangiogenesis after encephaloduroarteriosynangiosis in patients with moyamoya disease and type 2 diabetes mellitus: a case-control study.

OBJECTIVE: Diabetes is often linked to poorer outcomes in patients with moyamoya disease (MMD). However, experience has shown that certain individuals with diabetes have favorable outcomes after encephaloduroarteriosynangiosis (EDAS). The authors aimed to develop a nomogram to predict good neoangiogenesis in patients with MMD and type 2 diabetes mellitus (T2DM) to aid neurosurgeons in the identification of suitable candidates for EDAS. METHODS: Adults with MMD and T2DM who underwent EDAS between June 2004 and December 2018 were included in the analysis. In total, 126 patients (213 hemispheres) with MMD and T2DM from the Fifth Medical Centre of the Chinese PLA General Hospital were included and randomly divided into training (152 hemispheres) and internal validation (61 hemispheres) cohorts at a ratio of 7:3. Univariate logistic and least absolute shrinkage and selection operator regression analyses were used to identify the significant factors associated with good neoangiogenesis, which were used to develop a nomogram. The discrimination, calibration, and clinical utility were assessed. RESULTS: A total of 213 hemispheres in 126 patients were reviewed, including 152 (71.36%) hemispheres with good postoperative collateral formation and 61 (28.64%) with poor postoperative collateral formation. The authors selected 4 predictors (FGD5 rs11128722, VEGFA rs9472135, Suzuki stage, and internal carotid artery [ICA] moyamoya vessels) for nomogram development. The C-indices of the nomogram in the training and internal validation cohorts were 0.873 and 0.841, respectively. The nomogram exhibited a sensitivity of 84.5% and specificity of 81.0%. The positive and negative predictive values were 92.1% and 66.7%, respectively. The calibration curves indicated high predictive accuracy, and receiver operating characteristic curve analysis showed the superiority of the nomogram. The decision-making analysis validated the fitness and clinical application value of this nomogram. Then a web-based calculator to facilitate clinical application was generated. CONCLUSIONS: The nomogram developed in this study accurately predicted neoangiogenesis in patients with MMD and T2DM after EDAS and may assist neurosurgeons in identifying suitable candidates for indirect revascularization surgery.

Surface-Enhanced Raman Spectroscopy: Current Understanding, Challenges, and Opportunities.

While surface-enhanced Raman spectroscopy (SERS) has experienced substantial advancements since its discovery in the 1970s, it is an opportunity to celebrate achievements, consider ongoing endeavors, and anticipate the future trajectory of SERS. In this perspective, we encapsulate the latest breakthroughs in comprehending the electromagnetic enhancement mechanisms of SERS, and revisit CT mechanisms of semiconductors. We then summarize the strategies to improve sensitivity, selectivity, and reliability. After addressing experimental advancements, we comprehensively survey the progress on spectrum-structure correlation of SERS showcasing their important role in promoting SERS development. Finally, we anticipate forthcoming directions and opportunities, especially in deepening our insights into chemical or biological processes and establishing a clear spectrum-structure correlation.

Pyruvate kinase M2 sustains cardiac mitochondrial integrity in septic cardiomyopathy by regulating PHB2-dependent mitochondrial biogenesis.

Previous studies have highlighted the protective effects of pyruvate kinase M2 (PKM2) overexpression in septic cardiomyopathy. In our study, we utilized cardiomyocyte-specific PKM2 knockout mice to further investigate the role of PKM2 in attenuating LPS-induced myocardial dysfunction, focusing on mitochondrial biogenesis and prohibitin 2 (PHB2). Our findings confirmed that the deletion of PKM2 in cardiomyocytes significantly exacerbated LPS-induced myocardial dysfunction, as evidenced by impaired contractile function and relaxation. Additionally, the deletion of PKM2 intensified LPS-induced myocardial inflammation. At the molecular level, LPS triggered mitochondrial dysfunction, characterized by reduced ATP production, compromised mitochondrial respiratory complex I/III activities, and increased ROS production. Intriguingly, the absence of PKM2 further worsened LPS-induced mitochondrial damage. Our molecular investigations revealed that LPS disrupted mitochondrial biogenesis in cardiomyocytes, a disruption that was exacerbated by the absence of PKM2. Given that PHB2 is known as a downstream effector of PKM2, we employed PHB2 adenovirus to restore PHB2 levels. The overexpression of PHB2 normalized mitochondrial biogenesis, restored mitochondrial integrity, and promoted mitochondrial function. Overall, our results underscore the critical role of PKM2 in regulating the progression of septic cardiomyopathy. PKM2 deficiency impeded mitochondrial biogenesis, leading to compromised mitochondrial integrity, increased myocardial inflammation, and impaired cardiac function. The overexpression of PHB2 mitigated the deleterious effects of PKM2 deletion. This discovery offers a novel insight into the molecular mechanisms underlying septic cardiomyopathy and suggests potential therapeutic targets for intervention.

SYN1 variant causes X-linked neurodevelopmental disorders: a case report of variable clinical phenotypes in siblings.

The SYN1 gene encodes synapsin I, variants within the SYN1 gene are linked to X-linked neurodevelopmental disorders with high clinical heterogeneity, with reflex epilepsies (REs) being a representative clinical manifestation. This report analyzes a Chinese pedigree affected by seizures associated with SYN1 variants and explores the genotype-phenotype correlation. The proband, a 9-year-old boy, experienced seizures triggered by bathing at the age of 3, followed by recurrent absence seizures, behavioral issues, and learning difficulties. His elder brother exhibited a distinct clinical phenotype, experiencing sudden seizures during sleep at the age of 16, accompanied by hippocampal sclerosis. Whole exome sequencing (WES) confirmed a pathogenic SYN1 variant, c.1647_1650dup (p. Ser551Argfs*134), inherited in an X-linked manner from their mother. Notably, this variant displayed diverse clinical phenotypes in the two brothers and one previously reported case in the literature. Retrospective examination of SYN1 variants revealed an association between truncating variants and the pathogenicity of REs, and non-truncating variants are more related to developmental delay/intellectual disability (DD/ID). In summary, this study contributes to understanding complex neurodevelopmental disorders associated with SYN1, highlighting the clinical heterogeneity of gene variants and emphasizing the necessity for comprehensive genetic analysis in elucidating the pathogenic mechanisms of such diseases.

Signal2signal: Pushing the Spatiotemporal Resolution to the Limit by Single Chemical Hyperspectral Imaging.

There is growing interest in developing a high-performance self-supervised denoising algorithm for real-time chemical hyperspectral imaging. With a good understanding of the working function of the zero-shot Noise2Noise-based denoising algorithm, we developed a self-supervised Signal2Signal (S2S) algorithm for real-time denoising with a single chemical hyperspectral image. Owing to the accurate distinction and capture of the weak signal from the random fluctuating noise, S2S displays excellent denoising performance, even for the hyperspectral image with a spectral signal-to-noise ratio (SNR) as low as 1.12. Under this condition, both the image clarity and the spatial resolution could be significantly improved and present an almost identical pattern with a spectral SNR of 7.87. The feasibility of real-time denoising during imaging was well demonstrated, and S2S was applied to monitor the photoinduced exfoliation of transition metal dichalcogenide, which is hard to accomplish by confocal Raman spectroscopy. In general, the real-time denoising capability of S2S offers an easy way toward in situ/in vivo/operando research with much improved spatial and temporal resolution. S2S is open-source at https://github.com/3331822w/Signal2signal and will be accessible online at https://ramancloud.xmu.edu.cn/tutorial.

Double-Layered Hollow Mesoporous Cuprous Oxide Nanoparticles for Double Drug Sequential Therapy of Tumors.

Cancer stem cells (CSCs) are one of the determinants of tumor heterogeneity and are characterized by self-renewal, high tumorigenicity, invasiveness, and resistance to various therapies. To overcome the resistance of traditional tumor therapies resulting from CSCs, a strategy of double drug sequential therapy (DDST) for CSC-enriched tumors is proposed in this study and is realized utilizing the developed double-layered hollow mesoporous cuprous oxide nanoparticles (DL-HMCONs). The high drug-loading contents of camptothecin (CPT) and all-trans retinoic acid (ATRA) demonstrate that the DL-HMCON can be used as a generic drug delivery system. ATRA and CPT can be sequentially loaded in and released from CPT3@ATRA3@DL-HMCON@HA. The DDST mechanisms of CPT3@ATRA3@DL-HMCON@HA for CSC-containing tumors are demonstrated as follows: 1) the first release of ATRA from the outer layer induces differentiation from CSCs with high drug resistance to non-CSCs with low drug resistance; 2) the second release of CPT from the inner layer causes apoptosis of non-CSCs; and 3) the third release of Cu+ from DL-HMCON itself triggers the Fenton-like reaction and glutathione depletion, resulting in ferroptosis of non-CSCs. This CPT3@ATRA3@DL-HMCON@HA is verified to possess high DDST efficacy for CSC-enriched tumors with high biosafety.

Rapid Sample Pretreatment Facilitating SERS Detection of Trace Weak Organic Acids/Bases in Simple Matrices.

Surface-enhanced Raman spectroscopy (SERS) is a powerful tool for highly sensitive qualitative and quantitative analyses of trace targets. However, sensitive SERS detection can only be facilitated with a suitable sample pretreatment in fields related to trace amounts for food safety and clinical diagnosis. Currently, the sample pretreatment for SERS detection is normally borrowed and improved from the ones in the lab, which yields a high recovery but is tedious and time-consuming. Rapid detection of trace targets in a complex environment is still a considerable issue for SERS detection. Herein, we proposed a liquid-liquid extraction method coupled with a back-extraction method for sample pretreatment based on the pH-sensitive reversible phase transition of the weak organic acids and bases, where the lowest detectable concentrations were identical before and after the pretreatment process. The sensitive (µg L-1 level) and rapid (within 5 min) SERS detection of either koumine, a weak base, or celastrol, a weak acid, was demonstrated in different drinking water samples and beverages. Furthermore, target generality was demonstrated for a variety of weak acids and bases (2 < pKa < 12), and the hydrophilicity/hydrophobicity of the target determines the pretreatment efficiency. Therefore, the LLE-BE coupled SERS was developed as an easy, rapid, and low-cost tool for the trace detection of the two types of targets in simple matrices, which paved the way toward trace targets in complex matrices.

Deep Learning-Assisted Spectrum-Structure Correlation: State-of-the-Art and Perspectives.

Spectrum-structure correlation is playing an increasingly crucial role in spectral analysis and has undergone significant development in recent decades. With the advancement of spectrometers, the high-throughput detection triggers the explosive growth of spectral data, and the research extension from small molecules to biomolecules accompanies massive chemical space. Facing the evolving landscape of spectrum-structure correlation, conventional chemometrics becomes ill-equipped, and deep learning assisted chemometrics rapidly emerges as a flourishing approach with superior ability of extracting latent features and making precise predictions. In this review, the molecular and spectral representations and fundamental knowledge of deep learning are first introduced. We then summarize the development of how deep learning assist to establish the correlation between spectrum and molecular structure in the recent 5 years, by empowering spectral prediction (i.e., forward structure-spectrum correlation) and further enabling library matching and de novo molecular generation (i.e., inverse spectrum-structure correlation). Finally, we highlight the most important open issues persisted with corresponding potential solutions. With the fast development of deep learning, it is expected to see ultimate solution of establishing spectrum-structure correlation soon, which would trigger substantial development of various disciplines.

Structural Features, Physicochemical Properties, and In Vitro Digestibility of the Starch-Lipid Complexes Formed between High Amylose Starch and Stearic Acid or Potassium Stearate.

Starch-lipid complexes were prepared from high amylose starch (HAS) with stearic acid (SA) or potassium stearate (PS) at different molar concentrations. The complexes (HAS-PS) formed between HAS and PS showed polyelectrolyte characteristics with ζ-potential ranging from -22.2 to -32.8 mV, and the electrostatic repulsion between anionic charges restricted the starch chain reassociation and facilitated the formation of V-type crystalline structures upon cooling. The hydrophobic effects enabled recrystallization of the SA, and the HAS-SA complexes exhibited weaker V-type crystalline structures than the HAS-PS complexes; both HAS-SA/PS complexes were of a similar "mass fractal" type, with a dimension varied from 2.15 to 2.96. The HAS-SA complexes had a considerable content of resistant starch (RS, 16.1~29.2%), whereas negligible RS was found in the HAS-PS complexes. The findings from the present study imply that the molecular order of starch chains and the macro-structures of starch particles are more important to regulate the digestibility of starch-lipid complexes than the crystalline structures.

Leveraging molecular targeted drugs and immune checkpoint inhibitors treat advanced thyroid carcinoma to achieve thyroid carcinoma redifferentiation.

Thyroid cancer can be classified into three different types based on the degree of differentiation: well-differentiated, poorly differentiated, and anaplastic thyroid carcinoma. Well-differentiated thyroid cancer refers to cancer cells that closely resemble normal thyroid cells, while poorly differentiated and anaplastic thyroid carcinoma are characterized by cells that have lost their resemblance to normal thyroid cells. Advanced thyroid carcinoma, regardless of its degree of differentiation, is known to have a higher likelihood of disease progression and is generally associated with a poor prognosis. However, the process through which well-differentiated thyroid carcinoma transforms into anaplastic thyroid carcinoma, also known as "dedifferentiation", has been a subject of intensive research. In recent years, there have been significant breakthroughs in the treatment of refractory advanced thyroid cancer. Clinical studies have been conducted to evaluate the efficacy and safety of molecular targeted drugs and immune checkpoint inhibitors in the treatment of dedifferentiated thyroid cancer. These drugs work by targeting specific molecules or proteins in cancer cells to inhibit their growth or by enhancing the body's immune response against the cancer cells. This article aims to explore some of the possible mechanisms behind the dedifferentiation process in well-differentiated thyroid carcinoma. It also discusses the clinical effects of molecular targeted drugs and immune checkpoint inhibitors in thyroid cancer patients with different degrees of differentiation. Furthermore, it offers insights into the future trends in the treatment of advanced thyroid cancer, highlighting the potential for improved outcomes and better patient care.

Waterlogging in soil restricts the growth of Gleditsia sinensis seedlings and inhibits the accumulation of lignans and phenolic acids in thorns.

Gleditsia sinensis, commonly known as Chinese Zaojiao, has important economic value and medicinal compounds in its fruits and thorns, making it widely cultivated artificially in China. However, the available literature on the impact of waterlogging on the growth of G. sinensis seedlings and the accumulation of metabolite compounds in its thorns is limited. To address this knowledge gap, G. sinensis seedlings were planted in soil supplemented with pindstrup substrate, which enhances the water-holding capacity of the soil. The analyses of morphological traits and nutrient elements in one-year-old G. sinensis seedlings grown naturally under ambient conditions and metabolite accumulation in its thorns were conducted. The results showed that the waterlogged soil significantly diminished the height, fresh weight, and dry weight of seedling roots and stems (P < 0.05). Furthermore, waterlogging hindered the uptake of iron (Fe) and manganese (Mn), as well as the transport of potassium (K). The identified metabolites within the thorns were categorized into 16 distinct groups. Relative to the control soil, fatty acids and derivatives were the most down-regulated metabolites in the waterlogged soil, accounting for 40.58% of the total metabolites, followed by lignans (38.71%), phenolic acids (34.48%), saccharides and alcohols (34.15%), steroids (16.67%), alkaloids (12.24%), flavonoids (9.28%), and glycerophospholipids (7.41%). Conversely, nucleotides and derivatives experienced the greatest up-regulation in the waterlogged soil, accounting for 50.00% of the total metabolites. In conclusion, waterlogging negatively impacted the growth of G. sinensis seedlings and inhibited the accumulation of metabolites. Hence, when considering the accumulation of secondary metabolites such as lignans and phenolic acids, appropriate management of soil moisture levels should be taken into account.

A novel machine learning prediction model for metastasis in breast cancer.

BACKGROUND: Breast cancer (BC) metastasis is the common cause of high mortality. Conventional prognostic criteria cannot accurately predict the BC metastasis risk. The machine learning technologies can overcome the disadvantage of conventional models. AIM: We developed a model to predict BC metastasis using the random survival forest (RSF) method. METHODS: Based on demographic data and routine clinical data, we used RSF-recursive feature elimination to identify the predictive variables and developed a model to predict metastasis using RSF method. The area under the receiver operating characteristic curve (AUROC) and Kaplan-Meier survival (KM) analyses were plotted to validate the predictive effect when C-index was plotted to assess the discrimination and Brier scores was plotted to assess the calibration of the predictive model. RESULTS: We developed a metastasis prediction model comprising three variables (pathological stage, aspartate aminotransferase, and neutrophil count) selected by RSF-recursive feature elimination. The model was reliable and stable when assessed by the AUROC (0.932 in training set and 0.905 in validation set) and KM survival analyses (p < .0001). The C-indexes (0.959) and Brier score (0.097) also validated the good predictive ability of this model. CONCLUSIONS: This model relies on routine data and examination indicators in real-time clinical practice and exhibits an accurate prediction performance without increasing the cost for patients. Using this model, clinicians can facilitate risk communication and provide precise and efficient individualized therapy to patients with breast cancer.

MRI-Guided Tumor Therapy Based on Synergy of Ferroptosis, Immunosuppression Reversal and Disulfidptosis.

Triple negative breast cancer (TNBC) cells have a high demand for oxygen and glucose to fuel their growth and spread, shaping the tumor microenvironment (TME) that can lead to a weakened immune system by hypoxia and increased risk of metastasis. To disrupt this vicious circle and improve cancer therapeutic efficacy, a strategy is proposed with the synergy of ferroptosis, immunosuppression reversal and disulfidptosis. An intelligent nanomedicine GOx-IA@HMON@IO is successfully developed to realize this strategy. The Fe release behaviors indicate the glutathione (GSH)-responsive degradation of HMON. The results of titanium sulfate assay, electron spin resonance (ESR) spectra, 5,5'-Dithiobis-(2-nitrobenzoic acid (DTNB) assay and T1 -weighted magnetic resonance imaging (MRI) demonstrate the mechanism of the intelligent iron atom (IA)-based cascade reactions for GOx-IA@HMON@IO, generating robust reactive oxygen species (ROS). The results on cells and mice reinforce the synergistic mechanisms of ferroptosis, immunosuppression reversal and disulfidptosis triggered by the GOx-IA@HMON@IO with the following steps: 1) GSH peroxidase 4 (GPX4) depletion by disulfidptosis; 2) IA-based cascade reactions; 3) tumor hypoxia reversal; 4) immunosuppression reversal; 5) GPX4 depletion by immunotherapy. Based on the synergistic mechanisms of ferroptosis, immunosuppression reversal and disulfidptosis, the intelligent nanomedicine GOx-IA@HMON@IO can be used for MRI-guided tumor therapy with excellent biocompatibility and safety.

STGIC: A graph and image convolution-based method for spatial transcriptomic clustering.

Spatial transcriptomic (ST) clustering employs spatial and transcription information to group spots spatially coherent and transcriptionally similar together into the same spatial domain. Graph convolution network (GCN) and graph attention network (GAT), fed with spatial coordinates derived adjacency and transcription profile derived feature matrix are often used to solve the problem. Our proposed method STGIC (spatial transcriptomic clustering with graph and image convolution) is designed for techniques with regular lattices on chips. It utilizes an adaptive graph convolution (AGC) to get high quality pseudo-labels and then resorts to dilated convolution framework (DCF) for virtual image converted from gene expression information and spatial coordinates of spots. The dilation rates and kernel sizes are set appropriately and updating of weight values in the kernels is made to be subject to the spatial distance from the position of corresponding elements to kernel centers so that feature extraction of each spot is better guided by spatial distance to neighbor spots. Self-supervision realized by Kullback-Leibler (KL) divergence, spatial continuity loss and cross entropy calculated among spots with high confidence pseudo-labels make up the training objective of DCF. STGIC attains state-of-the-art (SOTA) clustering performance on the benchmark dataset of 10x Visium human dorsolateral prefrontal cortex (DLPFC). Besides, it's capable of depicting fine structures of other tissues from other species as well as guiding the identification of marker genes. Also, STGIC is expandable to Stereo-seq data with high spatial resolution.

Patch-Based Convolutional Encoder: A Deep Learning Algorithm for Spectral Classification Balancing the Local and Global Information.

Molecular vibrational spectroscopies, including infrared absorption and Raman scattering, provide molecular fingerprint information and are powerful tools for qualitative and quantitative analysis. They benefit from the recent development of deep-learning-based algorithms to improve the spectral, spatial, and temporal resolutions. Although a variety of deep-learning-based algorithms, including those to simultaneously extract the global and local spectral features, have been developed for spectral classification, the classification accuracy is still far from satisfactory when the difference becomes very subtle. Here, we developed a lightweight algorithm named patch-based convolutional encoder (PACE), which effectively improved the accuracy of spectral classification by extracting spectral features while balancing local and global information. The local information was captured well by segmenting the spectrum into patches with an appropriate patch size. The global information was extracted by constructing the correlation between different patches with depthwise separable convolutions. In the five open-source spectral data sets, PACE achieved a state-of-the-art performance. The more difficult the classification, the better the performance of PACE, compared with that of residual neural network (ResNet), vision transformer (ViT), and other commonly used deep learning algorithms. PACE helped improve the accuracy to 92.1% in Raman identification of pathogen-derived extracellular vesicles at different physiological states, which is much better than those of ResNet (85.1%) and ViT (86.0%). In general, the precise recognition and extraction of subtle differences offered by PACE are expected to facilitate vibrational spectroscopy to be a powerful tool toward revealing the relevant chemical reaction mechanisms in surface science or realizing the early diagnosis in life science.