A pathology foundation model for cancer diagnosis and prognosis prediction.

Histopathology image evaluation is indispensable for cancer diagnoses and subtype classification. Standard artificial intelligence methods for histopathology image analyses have focused on optimizing specialized models for each diagnostic task1,2. Although such methods have achieved some success, they often have limited generalizability to images generated by different digitization protocols or samples collected from different populations3. Here, to address this challenge, we devised the Clinical Histopathology Imaging Evaluation Foundation (CHIEF) model, a general-purpose weakly supervised machine learning framework to extract pathology imaging features for systematic cancer evaluation. CHIEF leverages two complementary pretraining methods to extract diverse pathology representations: unsupervised pretraining for tile-level feature identification and weakly supervised pretraining for whole-slide pattern recognition. We developed CHIEF using 60,530 whole-slide images spanning 19 anatomical sites. Through pretraining on 44 terabytes of high-resolution pathology imaging datasets, CHIEF extracted microscopic representations useful for cancer cell detection, tumour origin identification, molecular profile characterization and prognostic prediction. We successfully validated CHIEF using 19,491 whole-slide images from 32 independent slide sets collected from 24 hospitals and cohorts internationally. Overall, CHIEF outperformed the state-of-the-art deep learning methods by up to 36.1%, showing its ability to address domain shifts observed in samples from diverse populations and processed by different slide preparation methods. CHIEF provides a generalizable foundation for efficient digital pathology evaluation for patients with cancer.

Introducing enzymatic cleavage features and transfer learning realizes accurate peptide half-life prediction across species and organs.

Peptide drugs are becoming star drug agents with high efficiency and selectivity which open up new therapeutic avenues for various diseases. However, the sensitivity to hydrolase and the relatively short half-life have severely hindered their development. In this study, a new generation artificial intelligence-based system for accurate prediction of peptide half-life was proposed, which realized the half-life prediction of both natural and modified peptides and successfully bridged the evaluation possibility between two important species (human, mouse) and two organs (blood, intestine). To achieve this, enzymatic cleavage descriptors were integrated with traditional peptide descriptors to construct a better representation. Then, robust models with accurate performance were established by comparing traditional machine learning and transfer learning, systematically. Results indicated that enzymatic cleavage features could certainly enhance model performance. The deep learning model integrating transfer learning significantly improved predictive accuracy, achieving remarkable R2 values: 0.84 for natural peptides and 0.90 for modified peptides in human blood, 0.984 for natural peptides and 0.93 for modified peptides in mouse blood, and 0.94 for modified peptides in mouse intestine on the test set, respectively. These models not only successfully composed the above-mentioned system but also improved by approximately 15% in terms of correlation compared to related works. This study is expected to provide powerful solutions for peptide half-life evaluation and boost peptide drug development.

PPGR: a comprehensive perennial plant genomes and regulation database.

Perennial woody plants hold vital ecological significance, distinguished by their unique traits. While significant progress has been made in their genomic and functional studies, a major challenge persists: the absence of a comprehensive reference platform for collection, integration and in-depth analysis of the vast amount of data. Here, we present PPGR (Resource for Perennial Plant Genomes and Regulation; https://ngdc.cncb.ac.cn/ppgr/) to address this critical gap, by collecting, integrating, analyzing and visualizing genomic, gene regulation and functional data of perennial plants. PPGR currently includes 60 species, 847 million protein-protein/TF (transcription factor)-target interactions, 9016 transcriptome samples under various environmental conditions and genetic backgrounds. Noteworthy is the focus on genes that regulate wood production, seasonal dormancy, terpene biosynthesis and leaf senescence representing a wealth of information derived from experimental data, literature mining, public databases and genomic predictions. Furthermore, PPGR incorporates a range of multi-omics search and analysis tools to facilitate browsing and application of these extensive datasets. PPGR represents a comprehensive and high-quality resource for perennial plants, substantiated by an illustrative case study that demonstrates its capacity in unraveling gene functions and shedding light on potential regulatory processes.

Influence of topology on the phase transition of a ferromagnetic metal.

The topological ferromagnet CoS2 exhibits an anhysteretic, weakly first-order transition at the Curie temperature of 119.8 K with a tricritical point µ0Htcp at 0.034 T. Magnetic symmetry and the mixing of majority and minority spin eg bands at a subband crossing just above the Fermi level produce a topological component of the magnetization that leads to a negative M3 term in the Landau free energy. The position of the Fermi level relative to the subband crossing is critical for controlling the order of the transition. Hole doping in Co0.89Fe0.11S2 drains the minority-spin eg pocket and results in a normal second-order phase transition. Electron doping in Co0.94Ni0.06S2 raises the Fermi level toward the subband gap, producing a strongly first-order transition with 15 K hysteresis. Our results demonstrate a relation between topological electronic structure and thermal hysteresis at the Curie point, which may help in the search for magnetocaloric materials.

Bacterial cysteate dissimilatory pathway involves a racemase and d-cysteate sulfo-lyase.

The sulfite-reducing bacterium Bilophila wadsworthia, a common human intestinal pathobiont, is unique in its ability to metabolize a wide variety of sulfonates to generate sulfite as a terminal electron acceptor (TEA). The resulting formation of H2S is implicated in inflammation and colon cancer. l-cysteate, an oxidation product of l-cysteine, is among the sulfonates metabolized by B. wadsworthia, although the enzymes involved remain unknown. Here we report a pathway for l-cysteate dissimilation in B. wadsworthia RZATAU, involving isomerization of l-cysteate to d-cysteate by a cysteate racemase (BwCuyB), followed by cleavage into pyruvate, ammonia and sulfite by a d-cysteate sulfo-lyase (BwCuyA). The strong selectivity of BwCuyA for d-cysteate over l-cysteate was rationalized by protein structural modeling. A homolog of BwCuyA in the marine bacterium Silicibacter pomeroyi (SpCuyA) was previously reported to be a l-cysteate sulfo-lyase, but our experiments confirm that SpCuyA too displays a strong selectivity for d-cysteate. Growth of B. wadsworthia with cysteate as the electron acceptor is accompanied by production of H2S and induction of BwCuyA. Close homologs of BwCuyA and BwCuyB are present in diverse bacteria, including many sulfate- and sulfite-reducing bacteria, suggesting their involvement in cysteate degradation in different biological environments.

ELONGATED HYPOTCOTYL5 and SPINE BASE SIZE1 together mediate light-regulated spine expansion in cucumber.

Plant trichome development is influenced by diverse developmental and environmental signals, but the molecular mechanisms involved are not well understood in most plant species. Fruit spines (trichomes) are an important trait in cucumber (Cucumis sativus L.), as they affect both fruit smoothness and commercial quality. Spine Base Size1 (CsSBS1) has been identified as essential for regulating fruit spine size in cucumber. Here, we discovered that CsSBS1 controls a season-dependent phenotype of spine base size in wild-type plants. Decreased light intensity led to reduced expression of CsSBS1 and smaller spine base size in wild-type plants, but not in the mutants with CsSBS1 deletion. Additionally, knockout of CsSBS1 resulted in smaller fruit spine base size and eliminated the light-induced expansion of spines. Overexpression of CsSBS1 increased spine base size and rescued the decrease in spine base size under low light conditions. Further analysis revealed that ELONGATED HYPOTCOTYL5 (HY5), a major transcription factor involved in light signaling pathways, directly binds to the promoter of CsSBS1 and activates its expression. Knockout of CsHY5 led to smaller fruit spine base size and abolished the light-induced expansion of spines. Taken together, our study findings have clarified a CsHY5-CsSBS1 regulatory module that mediates light-regulated spine expansion in cucumber. This finding offers a strategy for cucumber breeders to develop fruit with stable appearance quality under changing light conditions.

METTL3-mediated m6A modification enhances lncRNA H19 stability to promote endothelial cell inflammation and pyroptosis to aggravate atherosclerosis.

This study explored the impact of N6-methyladenosine (m6A) modification on the regulation of long noncoding RNA (lncRNA) and atherosclerosis progression. An atherosclerosis cell model was established by treating human aortic endothelial cells (HAECs) with oxidized low-density lipoprotein. Additionally, an atherosclerotic animal model was developed using ApoE-/- C57BL/6 male mice fed a high-fat diet. Both models were employed to assess the expression changes of proteins associated with m6A modification. First, the effect of m6A modification writer protein methyltransferase-like 3 (METTL3) knockdown on changes in the level of pyroptosis in HAECs was investigated, and bioinformatic analysis confirmed that lncRNA H19 (H19) was the potential target of m6A modification. RNA-binding protein immunoprecipitation assays were subsequently performed to explore the interaction between H19 and the m6A writer protein METTL3, as well as the reader protein recombinant insulin-like growth factor 2 mRNA-binding protein 2 (IGF2BP2). Finally, the effect of H19 expression on pyroptosis levels in HAECs was evaluated. In the aortas of atherosclerosis mice, overall m6A levels were significantly elevated compared with controls (p < .05), with METTL3 and METTL14 mRNA and protein levels notably increased (p < .05). Similarly, ox-LDL-treated HAECs showed a significant rise in m6A levels, along with increased METTL3 and METTL14 expression (p < .05). METTL3 knockdown in HAECs led to decreased pyroptosis, as evidenced by reduced lactate dehydrogenase release and lower levels of IL-1ß, IL-18, and IL-6 (p < .05). Overexpression of H19 reversed these effects, indicating METTL3's role in promoting atherosclerosis by stabilizing H19 through m6A modification. H19 was the primary target lncRNA molecule of METTL3-mediated m6A modification in the pathogenesis of atherosclerosis. METTL3-mediated m6A modification regulated H19 expression, thereby aggravating atherosclerosis by activating pyroptosis.

Self-construction of actin networks through phase separation-induced abLIM1 condensates.

The abLIM1 is a nonerythroid actin-binding protein critical for stable plasma membrane-cortex interactions under mechanical tension. Its depletion by RNA interference results in sparse, poorly interconnected cortical actin networks and severe blebbing of migrating cells. Its isoforms, abLIM-L, abLIM-M, and abLIM-S, contain, respectively four, three, and no LIM domains, followed by a C terminus entirely homologous to erythroid cortex protein dematin. How abLIM1 functions, however, remains unclear. Here we show that abLIM1 is a liquid-liquid phase separation (LLPS)-dependent self-organizer of actin networks. Phase-separated condensates of abLIM-S-mimicking ΔLIM or the major isoform abLIM-M nucleated, flew along, and cross-linked together actin filaments (F-actin) to produce unique aster-like radial arrays and interconnected webs of F-actin bundles. Interestingly, ΔLIM condensates facilitated actin nucleation and network formation even in the absence of Mg2+. Our results suggest that abLIM1 functions as an LLPS-dependent actin nucleator and cross-linker and provide insights into how LLPS-induced condensates could self-construct intracellular architectures of high connectivity and plasticity.

Serine protease inhibitor E2 protects against cartilage tissue destruction and inflammation in osteoarthritis by targeting NF-κB signaling.

OBJECTIVE: Osteoarthritis (OA) is a chronic disease characterized by cartilage degeneration and inflammation, with no approved disease-modifying drugs. This study aimed to identify pathogenic genes and elucidate their mechanism in OA. METHODS: We systematically identified pathogenic genes combined sing-cell and bulk transcriptome profiles of cartilage tissues in OA. Adenovirus carrying the serpin peptidase inhibitor clade E member 2 (serpinE2) or exogenous serpinE2 was injected into monosodium iodoacetate (MIA)-induced OA-model rats. Histological analysis, immunohistochemistry, and Alcian blue staining were performed. In vitro, immunofluorescence, quantitative real-time PCR (RT-qPCR), enzyme-linked immunosorbent assay (ELISA), and western blot assays were performed. RESULTS: SerpinE2 exhibited elevated expression and hypomethylation, showing a positive association with collagen pathway activities in patients with OA. Silencing serpinE2 aggravated MIA-induced knee cartilage degeneration in OA-model rats. Conversely, the intra-articular injection of exogenous serpinE2 ameliorated articular cartilage degeneration, reduced pain-related behavioral responses, and relieve synovitis in MIA-induced OA-model rats. Exogenous serpinE2 not only attenuated the elevation of NLRP3, IL-1ß, and caspase1 expression levels but also restored the reduction in cell viability induced by lipopolysaccharide (LPS) in chondrocytes. Mechanistically, we found that exogenous serpinE2 inhibited LPS-induced reactive oxygen species (ROS) release and NF-κB signalling activation. CONCLUSIONS: SerpinE2 plays a protective role in cartilage and synovium tissues, suggesting that serpinE2 gene transfer or molecules that upregulate serpinE2 expression could be therapeutic candidates for OA.

Towards rapid colorimetric detection of extracellular vesicles using optofluidics-enhanced color-changing optical metasurface.

Efficient transportation and delivery of analytes to the surface of optical sensors are crucial for overcoming limitations in diffusion-limited transport and analyte sensing. In this study, we propose a novel approach that combines metasurface optics with optofluidics-enabled active transport of extracellular vesicles (EVs). By leveraging this combination, we show that we can rapidly capture EVs and detect their adsorption through a color change generated by a specially designed optical metasurface that produces structural colors. Our results demonstrate that the integration of optofluidics and metasurface optics enables spectrometer-less and label-free colorimetric read-out for EV concentrations as low as 107 EVs/ml, achieved within a short incubation time of two minutes.

Charge-transfer-driven ultrasensitive SERS sensing in a two-dimensional titanium carbonitride MXene.

Two-dimensional (2D) MXenes stand out as promising platforms for surface-enhanced Raman scattering (SERS) sensing owing to their metallic feature, various compositions, high surface area, compatibility with functionalization, and ease of fabrication. In this work, we report a high-performance 2D titanium carbonitride (Ti3CN) MXene SERS substrate. We reveal that the abundant electronic density of states near the Fermi level of Ti3CN MXene boosts the efficiency of photo-induced charge transfer at the interface of Ti3CN/molecule, resulting in significant Raman enhancement. The SERS sensitivity of Ti3CN MXene is further promoted through a 2D morphology regulation and molecular enrichment strategies. Moreover, prohibited drugs are detectable on this substrate, presenting the potential of trace-amount analysis on Ti3CN MXene. This work provides a deep insight of the SERS mechanisms of Ti3CN MXene and broadens the practical application of transition metal carbonitride MXene SERS substrates.

Photo/Mechanical/Acidic Multi-Stimuli Responses and Information Encryption Design of Acylhydrazone Derivative.

Stimuli-responsive crystalline materials have received much attention for being potential candidates of smart materials. However, the occurrence of polymorphism-driven stimuli responses in crystalline materials remains interesting but rare. Herein, three polymorphs of an acylhydrazone derivative, N'-[(E)-(1-benzofuran-2-yl) methylidene] pyridine -4-carbohydrazide (BFMP) were prepared. Form-1 undergoes a photomechanical response via E→Z photoisomerization under UV irradiation, accompanied by a decrease in fluorescence intensity and a change from colorless to yellow. Two types of Z→E thermal isomerization mechanisms with significant differences in conversion rate were observed at different temperatures in form-1. The solid-melt-solid transition has a faster conversion rate compared to the solid-solid transition due to freedom from lattice confinement. The transition from form-2 to form-3 can be achieved under grinding, coupled with a significant decrease in fluorescence intensity. The similar molecular stacking pattern of form-2 and form-3 provides a structural basis for the grinding-induced crystalline transition behavior. In addition, the presence of the pyridine moiety imparts an acidochromic property. The combination of photochromism and acidochromism explores the possible applications of acylhydrazone derivatives in information encryption.

Transition-Metal-Free Radical Relay Cascade Annulation of Amides: Access to Antitumor Active Benzo[b]azepine and Oxindole Derivatives.

Efficient transition-metal-free synthesis of benzo[b]azepines and oxindoles is achieved via a radical relay cascade strategy employing halogen atom transfer (XAT) for aryl radical generation followed by intramolecular hydrogen atom transfer (HAT). Optimization yielded moderate to substantial yields under visible light irradiation. Preliminary biological assessments revealed promising anti-tumor activity for select compounds. This study underscores the potential of XAT-mediated radical relay cascades in medicinal chemistry and anticancer drug discovery.

The prognostic role of pre-treatment neutrophil to lymphocyte ratio and platelet to lymphocyte ratio in esophageal squamous cell carcinoma treated with concurrent chemoradiotherapy.

PURPOSE: In this study, we retrospectively investigated the prognostic role of pre-treatment neutrophil to lymphocyte ratio (NLR) and platelet to lymphocyte ratio (PLR) in esophageal squamous cell carcinoma patients (ESCC) treated with concurrent chemo-radiotherapy (CCRT). METHODS: We retrospectively analyzed the records of 338 patients with pathologically diagnosed esophageal squamous cell carcinoma that underwent concurrent chemo-radiotherapy from January 2013 to December 2017. Univariate and multivariate analyses were used to identify prognostic factors for progression free survival (PFS) and overall survival (OS). RESULTS: The result showed that the thresholds for NLR and PLR were 2.47 and 136.0 by receiver operating characteristic curve. High NLR and PLR were both associated with tumor length (P < 0.05). High NLR and PLR were significantly associated with poor PFS and OS. Multivariate analyses identified NLR, PLR and TNM stage were independent risk factors for PFS and OS. CONCLUSIONS: We show that the pre-treatment NLR and PLR may serve as prognostic indicators for esophageal squamous cell carcinoma treated with concurrent chemo-radiotherapy.

HemoDL: Hemolytic peptides prediction by double ensemble engines from Rich sequence-derived and transformer-enhanced information.

Hemolytic peptides can trigger hemolysis by rupturing red blood cells' membranes and triggering cell disruption. Due to the labor-intensive and time-consuming in-lab identification process, accurate, high-throughput hemolytic peptide prediction is crucial for the growth of peptide sequence data in proteomics and peptidomics. In this study, we offer the HemoDL ensemble learning model, which learns the distinct distribution of sequence characteristics for predicting the hemolytic activity of peptides using a double LightGBM framework. To determine the most informative encoding features, we compare 17 widely used features across four benchmark datasets. Our investigation reveals that CTD, BPF, Charge, AAC, GDPC, ATC, QSO, and transformer-based features exhibit more positive contributions to detecting the hemolytic activity of peptides. Comparison with eight state-of-the-art methods demonstrates that HemoDL outperforms other models, attaining higher Matthews Correlation Coefficient values on four test datasets, ranging from 6.30% to 16.04%, 6.63%-11.26%, 4.76%-9.92%, and 7.41%-15.03%, respectively. Additionally, we provide the HemoDL with a user-friendly graphical interface available at https://github.com/abcair/HemoDL. In summary, the HemoDL model, leveraging CTD, BPF, Charge, AAC, GDPC, ATC, QSO and transformer-based encoding features within a double LightGBM learning framework, achieves high accuracy in predicting the hemolytic activity of peptides.

Allogeneic stem cell transplantation is still a highly curative therapy in adults with philadelphia chromosome-positive acute lymphoblastic leukaemia.

The application of tyrosine kinase inhibitors (TKIs) and novel immunotherapies has improved outcomes in patients with Ph + acute lymphoblastic leukaemia (ALL), and the issue of whether there is still a need for stem cell transplantation has become controversial. We performed a retrospective study to explore whether stem cell transplantation still held a place in patients with Ph + ALL if only imatinib and 2nd generation TKIs are available and affordable. A total of 292 patients were included. The median age was 38 years [range 14-64, IQR 28-48]. Patients receiving transplants (n = 216) had better rates of 4-year disease-free survival (DFS, 68% vs. 24%, P < .0001) and overall survival (OS, 72% vs. 47%, P < .0001) than those receiving continuous TKIs plus chemotherapy (TKI-chemo) (n = 76). In the multivariate analysis, male sex, WBC count ≥ 95 × 109/L and PLT count ≤ 154 × 109/L at diagnosis were significantly associated with poorer outcomes, and transplantation was significantly associated with favourable DFS and OS. In addition, the transplant outcomes were superior in any subgroup according to the number of risk variables. Furthermore, propensity score matching (PSM) analyses showed similar findings in the whole cohort and in age- and BCR-ABL1 level-based subgroups after the first or second consolidation. In conclusion, transplantation as a one-time procedure for adults with Ph + ALL patients remains important in countries lacking accessibility to third-generation TKIs or immunotherapies, regardless of the depth of the molecular response.

LLM4THP: a computing tool to identify tumor homing peptides by molecular and sequence representation of large language model based on two-layer ensemble model strategy.

Tumor homing peptides (THPs) have a distinctive capacity to specifically attach to tumor cells, providing a promising approach for targeted cancer treatment and detection. Although THPs have the potential for significant impact, their detection by conventional methods is both time-consuming and expensive. To tackle this issue, we provide LLM4THP, an innovative computational approach that utilizes large language models (LLMs) to quickly and effectively detect THPs. LLM4THP utilizes two protein LLMs, ESM2 and Prot_T5_XL_UniRef50, to encode peptide sequences. This allows for the capture of complex patterns and relationships within the peptide data. In addition, we utilize inherent sequence characteristics such as Amino Acid Composition (AAC), Pseudo Amino Acid Composition (PAAC), Amphiphilic Pseudo Amino Acid Composition (APAAC), and Composition, Transition, and Distribution (CTD) to improve the representation of peptides. The RDKitDescriptors feature representation approach transforms peptide sequences into molecular objects and computes chemical characteristics, resulting in enhanced THP identification. The LLM4THP ensemble strategy incorporates various features into a two-layer learning architecture. The first layer consists of LightGBM, XGBoost, Random Forest, and Extremely Randomized Trees, which generate a set of meta results. The second layer utilizes Logistic Regression to further refine the identification of sequences as either THP or non-THP. LLM4THP exhibits exceptional performance compared to the most advanced methods, showcasing enhancements in accuracy, Matthew's correlation coefficient, F1 score, area under the curve, and average precision. The source code and dataset can be accessed at the following URL: https://github.com/abcair/LLM4THP.

Surface mobility gradient and emergent facilitation in glassy films.

Confining glassy polymers into films can substantially modify their local and film-averaged properties. We present a lattice model of film geometry with void-mediated facilitation behaviors but free from any elasticity effect. We analyze the spatially varying viscosity to delineate the transport properties of glassy films. The film mobility measurements reported by Yang et al., Science, 2010, 328, 1676 are successfully reproduced. The flow exhibits a crossover from a simple viscous flow to a surface-dominated regime as the temperature decreases. The propagation of a highly mobile front induced by the free surface is visualized in real space. Our approach provides a microscopic treatment of the observed glassy phenomena.

Correction: Surface mobility gradient and emergent facilitation in glassy films.

Correction for 'Surface mobility gradient and emergent facilitation in glassy films' by Qiang Zhai et al., Soft Matter, 2024, https://doi.org/10.1039/D4SM00221K.