High-efficiency C3 electrosynthesis on a lattice-strain-stabilized nitrogen-doped Cu surface.

The synthesis of multi-carbon (C2+) fuels via electrocatalytic reduction of CO, H2O using renewable electricity, represents a significant stride in sustainable energy storage and carbon recycling. The foremost challenge in this field is the production of extended-chain carbon compounds (Cn, n ≥ 3), wherein elevated *CO coverage (θco) and its subsequent multiple-step coupling are both critical. Notwithstanding, there exists a "seesaw" dynamic between intensifying *CO adsorption to augment θco and surmounting the C-C coupling barrier, which have not been simultaneously realized within a singular catalyst yet. Here, we introduce a facilely synthesized lattice-strain-stabilized nitrogen-doped Cu (LSN-Cu) with abundant defect sites and robust nitrogen integration. The low-coordination sites enhance θco and concurrently, the compressive strain substantially fortifies nitrogen dopants on the catalyst surface, promoting C-C coupling activity. The n-propanol formation on the LSN-Cu electrode exhibits a 54% faradaic efficiency and a 29% half-cell energy efficiency. Moreover, within a membrane electrode assembly setup, a stable n-propanol electrosynthesis over 180 h at a total current density of 300 mA cm-2 is obtained.

Conversion chemoradiotherapy combined with nab-paclitaxel plus cisplatin in patients with locally advanced borderline-resectable or unresectable esophageal squamous cell carcinoma: a phase i/ii prospective cohort study.

BACKGROUND: To evaluate the efficacy and safety of nab-paclitaxel plus cisplatin as the regimen of conversional chemoradiotherapy (cCRT) in locally advanced borderline resectable or unresectable esophageal squamous cell carcinoma (ESCC). METHODS: Patients with locally advanced ESCC (cT3­4, Nany, M0­1, M1 was limited to lymph node metastasis in the supraclavicular area) were enrolled. All the patients received the cCRT of nab-paclitaxel plus cisplatin. After the cCRT, those resectable patients received esophagectomy; those unresectable patients continued to receive the definitive chemoradiotherapy (dCRT). The locoregional control (LRC), overall survival (OS), event-free survival (EFS), distant metastasis free survival (DMFS), pathological complete response (pCR), R0 resection rate, adverse events (AEs) and postoperative complications were calculated. RESULTS: 45 patients with ESCC treated from October 2019 to May 2021 were finally included. The median follow-up time was 30.3 months. The LRC, OS, EFS, DMFS at 1 and 2 years were 81.5%, 86.6%, 64.3%, 73.2 and 72.4%, 68.8%, 44.8%, 52.7% respectively. 21 patients (46.7%) received conversional chemoradiotherapy plus surgery (cCRT+S). The pCR rate and R0 resection rate were 47.6 and 84.0%. The LRC rate at 1 and 2 years were 95.0%, 87.1% in cCRT+S patitents and 69.3%, 58.7% in dCRT patients respectively (HR, 5.14; 95%CI, 1.10-23.94; P = 0.021). The toxicities during chemoradiotherapy were tolerated, and the most common grade 3-4 toxicitiy was radiation esophagitis (15.6%). The most common postoperative complication was pleural effusion (38.1%) and no grade ≥ IIIb complications were observed. CONCLUSION: nab-paclitaxel plus cisplatin are safe as the regimen of conversional chemoradiotherapy of ESCC.

ncRNA Coding Potential Prediction Using BiLSTM and Transformer Encoder-Based Model.

Many noncoding RNAs (ncRNAs) have been identified, and many of them play vital roles in various biological processes, including gene expression regulation, epigenetic regulation, transcription, and control. Recently, a few observations revealed that ncRNAs are translated into functional peptides. Moreover, many computational methods have been developed to predict the coding potential of these transcripts, which contributes to a deeper investigation of their functions. However, most of these are used to distinguish ncRNAs and mRNAs. It is important to develop a highly accurate computational tool for identifying the coding potential of ncRNAs, thereby contributing to the discovery of novel peptides. In this Article, we propose a novel BiLSTM And Transformer encoder-based model (nBAT) with intrinsic features encoded for ncRNA coding potential prediction. In nBAT, we introduce a learnable position encoding mechanism to better obtain the embeddings of the ncRNA sequence. Moreover, we extract 43 intrinsic features from different perspectives and encode these features into the Transformer encoder by calculating their distances. Our performance comparisons show that nBAT achieves a superior performance than the state-of-the-art methods for coding potential prediction on different datasets. We also apply the method to new ncRNAs for identifying the coding potential, and the results further indicate the competitive performance of nBAT. We expect the method can be exploited as a useful tool for high-throughput coding potential prediction for ncRNAs.

Higher immune cell radiation dose is correlated with poor tumor control and survival in patients with non-small cell lung cancer receiving postoperative radiotherapy.

INTRODUCTION: The estimated dose of radiation to immune cells (EDRIC) has been shown to correlate with the overall survival (OS) of patients who receive definitive thoracic radiotherapy. However, the planning target volume (PTV) may be a confounding factor. We assessed the prognostic value of EDRIC for non-small cell lung cancer (NSCLC) in patients who underwent postoperative radiotherapy (PORT) with homogeneous PTV. METHODS: Patients with NSCLC who underwent PORT between 2004 and 2019 were included. EDRIC was computed as a function of the number of radiation fractions and mean doses to the lungs, heart, and remaining body. The correlations between EDRIC and OS, disease-free survival (DFS), locoregional-free survival (LRFS), and distant metastasis-free survival (DMFS) were analyzed using univariate and multivariate Cox models. Kaplan-Meier analysis was performed to assess the survival difference between low- and high-EDRIC groups. RESULTS: In total, 345 patients were analyzed. The mean EDRIC was 6.26 Gy. Multivariate analysis showed that higher EDRIC was associated with worse outcomes in terms of OS (hazard ratio [HR] 1.207, P = .007), DFS (HR 1.129, P = .015), LRFS (HR 1.211, P = .002), and DMFS (HR 1.131, P = .057). In the low- and high-EDRIC groups, the 3-year OS was 81.2% and 74.0%, DFS 39.8% and 35.0%, LRFS 70.4% and 60.5%, and DMFS 73.9% and 63.1%, respectively. CONCLUSIONS: EDRIC is an independent prognostic factor for survival in patients with NSCLC undergoing PORT. Higher doses of radiation to the immune system are associated with tumor progression and poor survival. Organs at risk for the immune system should be considered during radiotherapy planning.

Disulfide-stabilized trimeric hemagglutinin ectodomains provide enhanced heterologous influenza protection.

Influenza virus infection poses a continual menace to public health. Here, we developed soluble trimeric HA ectodomain vaccines by establishing interprotomer disulfide bonds in the stem region, which effectively preserve the native antigenicity of stem epitopes. The stable trimeric H1 ectodomain proteins exhibited higher thermal stabilities in comparison with unmodified HAs and showed strong binding activities towards a panel of anti-stem cross-reactive antibodies that recognize either interprotomer or intraprotomer epitopes. Negative stain transmission electron microscopy (TEM) analysis revealed the stable trimer architecture of the interprotomer disulfide-stapled WA11#5, NC99#2, and FLD#1 proteins as well as the irregular aggregation of unmodified HA molecules. Immunizations of mice with those trimeric HA ectodomain vaccines formulated with incomplete Freund's adjuvant elicited significantly more potent cross-neutralizing antibody responses and offered broader immuno-protection against lethal infections with heterologous influenza strains compared to unmodified HA proteins. Additionally, the findings of our study indicate that elevated levels of HA stem-specific antibody responses correlate with strengthened cross-protections. Our design strategy has proven effective in trimerizing HA ectodomains derived from both influenza A and B viruses, thereby providing a valuable reference for designing future influenza HA immunogens.

Seamless rate adaptation for wide SNR range in FSO systems based on rate compatible modulation.

We propose and experimentally demonstrate a large-range and seamless rate-adaptive free-space optical (FSO) scheme based on rate compatible modulation (RCM). It utilizes a mapping method through weighted summation, enabling an even increase of bit energy. This allows the system to seamlessly adjust the throughput. In addition, the handover mechanism based on rate check and acknowledgment signals ensures a wide coverage of signal-to-noise ratio (SNR) fluctuation. It is worth noting that this scheme does not require any changes to the mapping matrix and decoding algorithm, making it easier to implement in real-world systems. Simulation proves the superiority of the proposed approach in terms of coding rate coverage by applying a fixed mapping matrix compared to a traditional adaptive modulation and coding scheme. Experimental demonstration over a 50 m FSO link verifies that the SNR dynamic range of this scheme is >15 dB with a seamless rate adjustment between 6.7 Gb/s and 53.6 Gb/s and the operability of RCM is not limited by the time scale of turbulence, revealing huge potential for the massive connections in future smart cities.

Low complexity sub-baud rate sampling reception in a bandwidth-limited IM/DD system.

This Letter presents what is to our knowledge a novel approach to reduce the digital signal processing (DSP) complexity in intensity modulation and direct detection (IM/DD) systems, which is critical for short-reach optical communication systems with severe bandwidth limitations. We propose a sub-baud rate sampling reception method utilizing a polyphase feedforward equalizer-based maximum likelihood sequence estimation (PFFE-MLSE), which could operate effectively under a sampling rate of 0.6 samples per symbol. This new architecture eliminates the need for resampling, allowing the adaptive equalizer to operate with significantly reduced complexity-over 60% compared to traditional FFE-MLSE. An offline experiment, transmitting a 100-Gbaud on-off keying (OOK) signal over a 5-km single-mode fiber (SMF) link, demonstrates the feasibility of our approach with bit error ratio (BER) meeting the KP4-forward error correction (KP4-FEC) threshold in the optical back-to-back (OBTB) scenario and 7% hard-decision FEC (HD-FEC) threshold in the 5-km SMF transmission.

Signal amplification strategies in photoelectrochemical sensing of carcinoembryonic antigen.

Early detection of cancer markers is critical for cancer diagnosis and cancer therapy since these markers may indicate cancer risk, incidence, and disease prognosis. Carcinoembryonic antigen (CEA) is a type of non-specific and broad-spectrum cancer biomarker commonly utilized for early cancer diagnosis. Moreover, it serves as an essential tool to assess the efficacy of cancer treatment and monitor tumor recurrence as well as metastasis, thus garnering significant attention for precise and sensitive CEA detection. In recent years, photoelectrochemical (PEC) techniques have emerged as prominent methods in CEA detection due to the advantages of PEC, such as simple equipment requirements, cost-effectiveness, high sensitivity, low interference from background signals, and easy of instrument miniaturization. Different signal amplification methods have been reported in PEC sensors for CEA analysis. Based on these, this article reviews PEC sensors based on various signal amplification strategies for detection of CEA during the last five years. The advantages and drawbacks of these sensors were discussed, as well as future challenges.

Hippocampal avoidance whole-brain radiotherapy with simultaneous integrated boost in lung cancer brain metastases and utility of the Hopkins verbal learning test for testing cognitive impairment in Chinese patients: a prospective phase II study.

BACKGROUND: This study aimed to evaluate the efficiency of hippocampal avoidance whole-brain radiotherapy with a simultaneous integrated boost (HA-WBRT-SIB) treating brain metastases (BM) and utility of the Hopkins Verbal Learning Test-Revised (HVLT-R) (Chinese version) in Chinese lung cancer patients. METHODS: Lung cancer patients with BM undergone HA-WBRT-SIB at our center were enrolled. Brain magnetic resonance imaging, The HVLT total learning score, and side effects were evaluated before radiotherapy and 1, 3, 6, and 12 months after radiotherapy. This study analyzed the overall survival rate, progression-free survival rate, and changes in HVLT-R immediate recall scores. RESULTS: Forty patients were enrolled between Jan 2016 and Jan 2020. The median follow-up time was 14.2 months. The median survival, progression-free survival, and intracranial progression-free survival of all patients were 14.8 months, 6.7 months and 14.8 months, respectively. Multivariate analysis indicated that male sex and newly diagnosed stage IV disease were associated with poor overall survival and progression-free survival, respectively. HVLT-R scores at baseline and 1, 3, and 6 months after radiotherapy were 21.94 ± 2.99, 20.88 ± 3.12, 20.03 ± 3.14, and 19.78 ± 2.98, respectively. The HVLT-R scores at 6 months after radiotherapy decreased by approximately 9.8% compared with those at baseline. No grade 3 toxicities occurred in the entire cohort. CONCLUSIONS: HA-WBRT-SIB is of efficiency and cognitive-conserving in treating Chinese lung cancer BM. TRIAL REGISTRATION: This study was retrospectively registered on ClinicalTrials.gov in 24th Feb, 2024. The ClinicalTrials.gov ID is NCT06289023.

TGC-ARG: Anticipating Antibiotic Resistance via Transformer-Based Modeling and Contrastive Learning.

In various domains, including everyday activities, agricultural practices, and medical treatments, the escalating challenge of antibiotic resistance poses a significant concern. Traditional approaches to studying antibiotic resistance genes (ARGs) often require substantial time and effort and are limited in accuracy. Moreover, the decentralized nature of existing data repositories complicates comprehensive analysis of antibiotic resistance gene sequences. In this study, we introduce a novel computational framework named TGC-ARG designed to predict potential ARGs. This framework takes protein sequences as input, utilizes SCRATCH-1D for protein secondary structure prediction, and employs feature extraction techniques to derive distinctive features from both sequence and structural data. Subsequently, a Siamese network is employed to foster a contrastive learning environment, enhancing the model's ability to effectively represent the data. Finally, a multi-layer perceptron (MLP) integrates and processes sequence embeddings alongside predicted secondary structure embeddings to forecast ARG presence. To evaluate our approach, we curated a pioneering open dataset termed ARSS (Antibiotic Resistance Sequence Statistics). Comprehensive comparative experiments demonstrate that our method surpasses current state-of-the-art methodologies. Additionally, through detailed case studies, we illustrate the efficacy of our approach in predicting potential ARGs.

Ensemble Machine Learning and Predicted Properties Promote Antimicrobial Peptide Identification.

The emergence of antibiotic-resistant microbes raises a pressing demand for novel alternative treatments. One promising alternative is the antimicrobial peptides (AMPs), a class of innate immunity mediators within the therapeutic peptide realm. AMPs offer salient advantages such as high specificity, cost-effective synthesis, and reduced toxicity. Although some computational methodologies have been proposed to identify potential AMPs with the rapid development of artificial intelligence techniques, there is still ample room to improve their performance. This study proposes a predictive framework which ensembles deep learning and statistical learning methods to screen peptides with antimicrobial activity. We integrate multiple LightGBM classifiers and convolution neural networks which leverages various predicted sequential, structural and physicochemical properties from their residue sequences extracted by diverse machine learning paradigms. Comparative experiments exhibit that our method outperforms other state-of-the-art approaches on an independent test dataset, in terms of representative capability measures. Besides, we analyse the discrimination quality under different varieties of attribute information and it reveals that combination of multiple features could improve prediction. In addition, a case study is carried out to illustrate the exemplary favorable identification effect. We establish a web application at http://amp.denglab.org to provide convenient usage of our proposal and make the predictive framework, source code, and datasets publicly accessible at https://github.com/researchprotein/amp .

Preclinical study and phase II trial of adapting low-dose radiotherapy to immunotherapy in small cell lung cancer.

BACKGROUND: Immune checkpoint inhibitors (ICIs) provide modest but unsatisfactory benefits for extensive-stage small cell lung cancer (ES-SCLC). Developing strategies for treating ES-SCLC is critical. METHODS: We preliminarily explored the outcomes of salvage low-dose radiotherapy (LDRT) plus ICI on refractory SCLC patients. Next, we evaluated the combinational efficacy in murine SCLC. The tumor immune microenvironment (TIME) was analyzed for mechanistic study. Subsequently, we conducted a multicenter, prospective phase II trial that administered concurrent thoracic LDRT plus chemoimmunotherapy to treatment-naive ES-SCLC patients (MATCH trial, NCT04622228). The primary endpoint was confirmed objective response rate (ORR), and the key secondary endpoints included progression-free survival (PFS) and safety. FINDINGS: Fifteen refractory SCLC patients treated with LDRT plus ICI were retrospectively reviewed. The ORR was 73.3% (95% confidence interval [CI], 44.9-92.2). We identified a specific dose of LDRT (15 Gy/5 fractions) that exhibited growth retardation and improved survival in murine SCLC when combined with ICIs. This combination recruited a special T cell population, TCF1+ PD-1+ CD8+ stem-like T cells, from tumor-draining lymph nodes into the TIME. The MATCH trial showed a confirmed ORR of 87.5% (95% CI, 75.9-94.8). The median PFS was 6.9 months (95% CI, 5.4-9.3). CONCLUSIONS: These findings verified that LDRT plus chemoimmunotherapy was safe, feasible, and effective for ES-SCLC, warranting further investigation. FUNDING: This research was funded by West China Hospital (no. ZYJC21003), the National Natural Science Foundation of China (no. 82073336), and the MATCH trial was fully funded by Roche (China) Holding Ltd. (RCHL) and Shanghai Roche Pharmaceuticals Ltd. (SRPL).

Effectiveness and safety of combined nimotuzumab and S-1 chemotherapy with concurrent radiotherapy for locally advanced esophageal cancer in malnourished and elderly patients: A prospective phase II study.

Objective: Definitive chemoradiotherapy (dCRT) is the standard treatment for unresectable locally advanced esophageal cancer. However, this treatment is associated with substantial toxicity, and most malnourished or elderly patients are unable to complete this therapy. Therefore, there is a need for a more suitable radiotherapy combination regimen for this population. This study was aimed to evaluate the efficacy and safety of a combination regimen comprising chemotherapy with nimotuzumab and S-1 and concurrent radiotherapy for patients with fragile locally advanced esophageal cancer with a high Nutritional Risk Screening 2002 (NRS-2002) score. Methods: Eligible patients with unresectable esophageal carcinoma who had an NRS-2002 score of 2 or higher were enrolled. They were treated with S-1 and nimotuzumab with concurrent radiotherapy, followed by surgery or definitive radiotherapy. The primary endpoint was the locoregional control (LRC) rate. Results: A total of 55 patients who met the study criteria were enrolled. After completion of treatment, surgery was performed in 15 patients and radiotherapy was continued in 40 patients. The median follow-up period was 33.3 [95% confidence interval (95% CI), 31.4-35.1)] months. The LRC rate was 77.2% (95% CI, 66.6%-89.4%) at 1 year in the entire population. The overall survival (OS) rate and event-free survival (EFS) rate were 57.5% and 51.5% at 3 years, respectively. Surgery was associated with better LRC [hazard ratio (HR)=0.16; 95% CI, 0.04-0.70; P=0.015], OS (HR=0.19; 95% CI, 0.04-0.80; P=0.024), and EFS (HR=0.25; 95% CI, 0.08-0.75; P=0.013). Most adverse events were of grade 1 or 2, and no severe adverse events occurred. Conclusions: For malnourished or elderly patients with locally advanced esophageal cancer, radiotherapy combined with nimotuzumab and S-1 is effective and has a good safety profile.

Joint association of sleep quality and physical activity with metabolic dysfunction-associated fatty liver disease: a population-based cross-sectional study in Western China.

BACKGROUND: Metabolic dysfunction-associated fatty liver disease (MAFLD) is a growing threat leading to substantial disease burden globally. Poor sleep and physical inactivity are common in modern societies and independently associated with MAFLD, however, their joint effects on MAFLD remains unclear. METHODS: This population-based cross-sectional study was conducted in Xinjiang Uygur Autonomous Region, China, between July 2019 and September 2021. Self-reported sleep behaviors and physical activity (PA) were assessed using validated questionnaires. The primary outcome was radiological diagnosis of MAFLD. RESULTS: Of the 10 089 participants aged 47.0 (9.1) years (51.6% men), 3854 (38.2%) individuals had MAFLD. Poor sleep quality and physical inactivity were independently and jointly associated with an increased prevalence of MAFLD, independent of traditional risk factors (P < 0.05). Compared to subjects with guideline-recommended moderate-to-vigorous PA (MVPA) and good sleep quality, individuals with no recommended MVPA and poor sleep had the highest possibility of MAFLD (odds ratio = 2.36, 95% confidence interval: 1.81 - 3.08). Enhancing sleep quality substantially attenuated MAFLD prevalence regardless of the volume of PA, whereas, engaging in PA well above current guidelines did not adequately counteract the adverse impacts of poor sleep on MAFLD. CONCLUSIONS: Public health awareness and strategies concurrently targeting both sleep quality and PA should be encouraged to curb the climbing prevalence of MAFLD.

The impact of vegetation reconstruction on soil erosion in the Loess plateau.

Vegetation restoration not only extensively reshapes spatial land use patterns but also profoundly affects the dynamics of runoff and sediment loss. However, the influence of vegetation restoration on runoff and sediment yield from a regional perspective are scarce. This study therefore focused on 85 sites within the "Grain for Green" Project (GGP) region on the Loess Plateau, to investigate the impacts of the GGP on soil erosion. The results revealed a notable reduction in sediment loss and runoff due to vegetation restoration. Since the inception of the GGP in 1999, approximately 4.1 × 106 ha of degraded lands have been converted into forestlands, shrublands, and grasslands, resulting in an average annual reduction of 1.4 × 109 m3 in runoff and a decrease of 3.6 × 108 t in annual sediment loss on the whole Loess Plateau, with the GGP contributing approximately 26.7% of the sediment reduction in the Yellow River basin. The reduced soil erosion has mainly been regulated by vegetation cover, soil properties (clay, silt, and sand), slope, and precipitation on the Loess Plateau. The insights gained offer valuable contributions to large-scale assessments of changes in soil erosion in response to vegetation reconstruction and enhance our understanding of the spatial configurations associated with soil erosion control measures.

Efficacy and risk of donor-derived CAR-T treatment of relapsed B-cell acute lymphoblastic leukemia after hematopoietic stem cell transplantation.

The one-year survival rate for patients experiencing a relapse of B-cell acute lymphocytic leukemia (B-ALL) following hematopoietic stem cell transplantation (HSCT) is approximately 30%. Patients experiencing a relapse after allogeneic HSCT frequently encounter difficulties in obtaining autologous CAR-T products. We conducted a study involving 14 patients who received donor-derived CAR-T therapy for relapsed B-ALL following HSCT between August 2019 and May 2023 in our center. The results revealed a CR/CRi rate of 78.6% (11/14), a GVHD rate of 21.4% (3/14), and a 1-year overall survival (OS) rate of 56%. Decreased bone marrow donor cell chimerism in 9 patients recovered after CAR-T therapy. The main causes of death were disease progression and infection. Further analysis showed that GVHD (HR 7.224, 95% CI 1.42-36.82, P = 0.017) and platelet recovery at 30 days (HR 6.807, 95% CI 1.61-28.83, P = 0.009) are significantly associated with OS after CAR-T therapy. Based on the findings, we conclude that donor-derived CAR-T cells are effective in treating relapsed B-ALL patients following HSCT. Additionally, GVHD and poor platelet recovery impact OS, but further verification with a larger sample size is needed.

Robust and adaptive star identification algorithm based on linear assignment for multiple large field of view visual imaging systems.

The integration of the visual imaging system and the self-attitude determination system in on-orbit space projects necessitates robust star identification algorithms. However, disturbances in the on-orbit environment pose a challenge to the accuracy and efficiency of star identification algorithms. This paper introduces a novel star identification algorithm, to the best of our knowledge, designed for multiple large field of view (FOV) visual imaging systems, providing stability in the presence of the noise types, including position noise, lost-star noise, and fake-star noise. We employ the dynamic simulated star images generation method to construct simulated star image libraries suitable for various cameras with different FOV angles. Our algorithm comprises two parts: the star edge matching for coarse matching of interstellar angular distances based on linear assignment, and the star point registration for precise matching of star vectors. This innovative combination of local edge generation and global matching approach achieves an impressive 97.83% identification accuracy, maintaining this performance even with a standard deviation of one pixel in image plane errors and up to five missing stars. A comprehensive approach involving both simulated and empirical experiments was employed to validate the algorithm's effectiveness. This integration of the visual imaging system and the self-attitude determination system offers potential benefits such as reduced space equipment weight, simplified satellite launch processes, and decreased maintenance costs.

Chromosome-level genome assembly of Solanum pimpinellifolium.

Solanum pimpinellifolium, the closest wild relative of the domesticated tomato, has high potential for use in breeding programs aimed at developing multi-pathogen resistance and quality improvement. We generated a chromosome-level genome assembly of S. pimpinellifolium LA1589, with a size of 833 Mb and a contig N50 of 31 Mb. We anchored 98.80% of the contigs into 12 pseudo-chromosomes, and identified 74.47% of the sequences as repetitive sequences. The genome evaluation revealed BUSCO and LAI score of 98.3% and 14.49, respectively, indicating high quality of this assembly. A total of 41,449 protein-coding genes were predicted in the genome, of which 89.17% were functionally annotated. This high-quality genome assembly serves as a valuable resource for accelerating the biological discovery and molecular breeding of this important horticultural crop.

MvMRL: a multi-view molecular representation learning method for molecular property prediction.

Effective molecular representation learning is very important for Artificial Intelligence-driven Drug Design because it affects the accuracy and efficiency of molecular property prediction and other molecular modeling relevant tasks. However, previous molecular representation learning studies often suffer from limitations, such as over-reliance on a single molecular representation, failure to fully capture both local and global information in molecular structure, and ineffective integration of multiscale features from different molecular representations. These limitations restrict the complete and accurate representation of molecular structure and properties, ultimately impacting the accuracy of predicting molecular properties. To this end, we propose a novel multi-view molecular representation learning method called MvMRL, which can incorporate feature information from multiple molecular representations and capture both local and global information from different views well, thus improving molecular property prediction. Specifically, MvMRL consists of four parts: a multiscale CNN-SE Simplified Molecular Input Line Entry System (SMILES) learning component and a multiscale Graph Neural Network encoder to extract local feature information and global feature information from the SMILES view and the molecular graph view, respectively; a Multi-Layer Perceptron network to capture complex non-linear relationship features from the molecular fingerprint view; and a dual cross-attention component to fuse feature information on the multi-views deeply for predicting molecular properties. We evaluate the performance of MvMRL on 11 benchmark datasets, and experimental results show that MvMRL outperforms state-of-the-art methods, indicating its rationality and effectiveness in molecular property prediction. The source code of MvMRL was released in https://github.com/jedison-github/MvMRL.