Multi-omics analysis of immune-related microbiome and prognostic model in head and neck squamous cell carcinoma.

OBJECTIVES: The aim of our study is to explore the transcriptional and microbial characteristics of head and neck cancer's immune phenotypes using a multi-omics approach. MATERIALS AND METHODS: Employing TCGA data, we analyzed head and neck squamous cell carcinoma (HNSCC) immune cells with CIBERSORT and identified differentially expressed genes using DESeq2. Microbial profiles, obtained from the TCMA database, were analyzed using LEfSe algorithm to identify differential microbes in immune cell infiltration (ICI) subgroups. Random Forest algorithm and deep neural network (DNN) were employed to select microbial features and developed a prognosis model. RESULTS: We categorized HNSCC into three immune subtypes, finding ICI-2 with the worst prognosis and distinct microbial diversity. Our immune-related microbiome (IRM) model outperformed the TNM staging model in predicting survival, linking higher IRM model scores with poorer prognosis, and demonstrating clinical utility over TNM staging. Patients categorized as low-risk by the IRM model showed higher sensitivity to cisplatin and sorafenib treatments. CONCLUSIONS: This study offers a comprehensive exploration of the ICI landscape in HNSCC. We provide a detailed scenario of immune regulation in HNSCC and report a correlation between differing ICI patterns, intratumor microbiome, and prognosis. This research aids in identifying prime candidates for optimizing treatment strategies in HNSCC. CLINICAL RELEVANCE: This study revealed the microbial signatures associated with immunophenotyping of HNSCC and further found the microbial signatures associated with prognosis. The prognostic model based on IRM microbes is helpful for early prediction of patient prognosis and assisting clinical decision-making.

Genesis and Accumulation Mechanism of Low-Rank CBM: A Case Study in the Jiergalangtu Block of Erlian Basin, Northern China.

In order to elucidate the origin of coalbed methane (CBM) in the Jiergalangtu block of Erlian Basin, Inner Mongolia of China, gas components, stable isotope tests of 22 gas samples, radioisotope dating measurements, and water quality analysis of 15 coproduced water samples were evaluated. On account of the geochemical data and genetic indicators, including C1/C1-n, C1/(C2 + C3), and CO2/(CO2 + CH4) (CDMI) values, δ13C(CO2), Δδ13C(CO2-CH4), δ15N, and 3He/4He combined with vitrinite reflectance (Ro) (0.29-0.48%, avg. 0.35%) of Saihantala formation, the results indicate that methane in the Jiergalangtu block is mostly dominated by primary and secondary biological gas, 40.91% of the gas samples are secondary biogas and primary biogas accounts for 59.19%. Among them, methyl-type fermentation accounts for 31.82%, and carbon dioxide (CO2) reduction makes up 68.18%. CO2 reduction generally occurs region-wide but is mainly associated with the central part of the block, where CO2 depletion and 13C enrichment take place correspondingly. Methane and CO2 δ13C almost tend to isotopically light along the margin of the block, indicating that gas generation is significantly affected by the methyl-type fermentation pathway. Meanwhile, the genesis analysis of other gas components in CBM is also investigated, CO2 is mainly the associated product of microbial methanogenesis, and nitrogen (N2) is primarily from the atmosphere with a little amount from the earth's crust. Furthermore, the formation time of coalbed water has been dissected based on the hydrogeochemical properties of the coproduced water samples. The coalbed water exhibit a Na-HCO3 and Na-HCO3-Cl type and have a total dissolved solid (TDS) value ranging from 2458.58 to 5579.1 mg/L, with an average of 3440.55 mg/L. Moreover, comprehensive analysis of δD(H2O), δ18O(H2O), δ13CDIC, and the radioisotope dating index [3H, 14C(Fm) and 14C(BP)] indicates that the coalbed water was formed in the Quaternary Pleistocene and rarely replenished by the present surface water. The mechanism of CBM accumulation is basically sorted out by synthesizing the history of burial, heat, and hydrocarbon generation. The CBM formation can be divided into four stages. That is, microbial gas production approximately began at the beginning of the Early Cretaceous and reached the peak of thermogenic gas production in the middle and late Early Cretaceous. At the end of the Early Cretaceous, strata possibly began to uplift, and denudation led to gas escape. From Neogene to Pleistocene, glacial meltwater tended to penetrate into coalbed on a large scale, and N2 and CO2 also entered the coal seams, stimulating abundant secondary biological gas generation. Since Holocene, geological conditions including temperature and TDS have become hostile to biogas generation, and biogas generation tends to stop. Therefore, the Jiergalangtu block mainly represents sealed primary biological gas and secondary biological gas in CBM reservoirs.

Epithelial SIRT6 governs IL-17A pathogenicity and drives allergic airway inflammation and remodeling.

Dysregulation of IL-17A is closely associated with airway inflammation and remodeling in severe asthma. However, the molecular mechanisms by which IL-17A is regulated remain unclear. Here we identify epithelial sirtuin 6 (SIRT6) as an epigenetic regulator that governs IL-17A pathogenicity in severe asthma. Mice with airway epithelial cell-specific deletion of Sirt6 are protected against allergen-induced airway inflammation and remodeling via inhibiting IL-17A-mediated inflammatory chemokines and mesenchymal reprogramming. Mechanistically, SIRT6 directly interacts with RORγt and mediates RORγt deacetylation at lysine 192 via its PPXY motifs. SIRT6 promotes RORγt recruitment to the IL-17A gene promoter and enhances its transcription. In severe asthma patients, high expression of SIRT6 positively correlates with airway remodeling and disease severity. SIRT6 inhibitor (OSS_128167) treatment significantly attenuates airway inflammation and remodeling in mice. Collectively, these results uncover a function for SIRT6 in regulating IL-17A pathogenicity in severe asthma, implicating SIRT6 as a potential therapeutic target for severe asthma.

Assessment of microplastic contamination in an eastern Pacific tuna (Katsuwonus pelamis) and evaluation of its health risk implication through molecular docking and metabolomics studies.

This work investigated microplastic (MP) pollution in a commercially-important tuna species Katsuwonus pelamis (K. pelamis) from the Eastern Pacific and health implications. 125 MPs were extracted from gills, esophagus, stomachs, intestinal tracts, and muscle of K. pelamis. MPs in the esophagus was the highest, ∼7.6 times higher than that in the gill. Polyester and polyethylene terephthalate (PET) were dominant. Molecular docking implied that PET stabilized the complex via forming 4 new hydrogen bonds that interacted with Arg83, Gln246, Thr267, and Gly268, given that PET can enter glycerol kinase protein active pocket. Metabonomic results suggested that Glycerol 3-phosphate up expressed 1.66 more times that of control groups with no MPs in the muscle. This confirmed that MPs would lie in the glycerol kinase protein active pocket, which triggered menace to K. pelamis. The results provided insights into suggested the potential influence of MPs on the sustainability of fisheries and seafood safety.

The Unique Genetic Mutation Characteristics Based on Large Panel Next-Generation Sequencing (NGS) Detection in Multiple Primary Lung Cancers (MPLC) Patients.

BACKGROUND: With the wide application of multislice spiral computed tomography (CT), the frequency of detection of multiple lung cancer is increasing. This study aimed to analyze gene mutations characteristics in multiple primary lung cancers (MPLC) using large panel next-generation sequencing (NGS) assays. METHODS: Patients with MPLC surgically removed from the Affiliated Hospital of Guangdong Medical University from Jan 2020 to Dec 2021 enrolled the study. NGS sequencing of large panels of 425 tumor-associated genes was performed. RESULTS: The 425 panel sequencing of 114 nodules in 36 patients showed that epidermal growth factor receptor (EGFR) accounted for the largest proportion (55.3%), followed by Erb-B2 Receptor Tyrosine Kinase 2 (ERBB2) (9.6%), v-Raf murine sarcoma viral oncogene homolog B1 (BRAF), and Kirsten rat sarcoma viral oncogene (KRAS) (8.8%). Fusion target variation was rare (only 2, 1.8%). ERBB2 Y772_A775dup accounted for 73%, KRAS G12C for about 18%, and BRAF V600E for only 10%. AT-rich interaction domain 1A (ARID1A) mutations were significantly higher in invasive adenocarcinoma (IA) which contained solid/micro-papillary malignant components (p = 0.008). The tumor mutation burden (TMB) distribution was low, with a median TMB of 1.1 MUTS/Mb. There were no differences in the TMB distribution of different driver genes. In addition, 97.2% of MPLC patients (35/36) had driver gene mutations, and 47% had co-mutations, mainly in IA (45%) and invasive adenocarcinoma (MIA) (37%) nodule, with EGFR (39.4%), KRAS (9.1%), ERBB2 (6.1%), tumor protein 53 (TP53) (6.1%) predominately. CONCLUSIONS: MPLC has a unique genetic mutation characteristic that differs from advanced patients and usually presents with low TMB. Comprehensive NGS helps to diagnose MPLC and guides the MPLC clinical treatment. ARID1A is significantly enriched in IA nodules containing micro-papillary/solid components, suggesting that these MPLC patients may have a poor prognosis.

An innovative biomass-driven energy systems for green energy and freshwater production with less CO2 emission: Environmental and technical approaches.

Considering the current crisis of fossil energies, the exploitation of renewables and green technologies is necessary and unavoidable. Additionally, the design and development of integrated energy systems with two or more output products and the maximum usage of thermal losses in order to improve efficiency can boost the yield and acceptability of the energy system. In this regard, this paper develops a comprehensive multi-aspect assessment of the operation of a new solar and biomass energies-driven multigeneration system (MGS). The main units installed in MGS are three electric energy generation units based on a gas turbine process, a solid oxide fuel cell unit (SOFCU) and an organic Rankine cycle unit (ORCU), a biomass energy conversion unit to useful thermal energy, a seawater conversion unit into useable freshwater, a unit for converting water and electricity into hydrogen energy and oxygen gas, a unit for converting solar energy into useful thermal energy (based on Fresnel collector), and a cooling load generation unit. The planned MGS has a novel configuration and layout that has not been considered by researchers recently. The current article is based on presenting a multi-aspect evaluation to study thermodynamic-conceptual, environmental and exergoeconomic analyzes. The outcomes indicated that the planned MGS can produce about 6.31 MW of electrical power and 0.49 MW of thermal power. Furthermore, MGS is able to produce various products such as potable water (∼0.977 kg/s), cooling load (∼0.16 MW), hydrogen energy (∼1.578 g/s) and sanitary water (∼0.957 kg/s). The total thermodynamic indexes were calculated as 78.13% and 47.72%, respectively. Also, the total investment and unit exergy costs were 47.16 USD per hour and 11.07 USD per GJ, respectively. Further, the content of CO2 emitted from the desgined system was equal to 10.59 kmol per MWh. A parametric study has been also developed to identify influencing parameters.

Simulation of CO2 capture from natural gas by cyclic pressure swing adsorption process using activated carbon.

This work presented modeling and simulation of CO2 from natural gas. One of the most promising technologies is Pressure Swing Adsorption (PSA), which is an energy-efficient and cost-effective process for separating and capturing CO2 from industrial processes and power plants. This paper provides an overview of the PSA process and its application for CO2 capture, along with a discussion of its advantages, limitations, and future research directions. This process is pressure swing adsorption (PSA) with four adsorption beds. The adsorption bed columns fill with activated carbon as adsorbent. In this simulation momentum, mass and energy balance are solved simultaneously. The process was designed with two beds in adsorption conditions and the other two beds in desorption conditions. The desorption cycle includes blow-down and purge steps. The linear driving force (LDF) estimates the adsorption rate in modeling this process. The extended Langmuir isotherm is used for the equilibrium between solid and gas phases. The temperature changes by heat transfer from the gas phase to solid and axial heat dispersion. The set of partial differential equations is solved using implicit finite difference.

Homophily-Enhanced Self-Supervision for Graph Structure Learning: Insights and Directions.

Graph neural networks (GNNs) have recently achieved remarkable success on a variety of graph-related tasks, while such success relies heavily on a given graph structure that may not always be available in real-world applications. To address this problem, graph structure learning (GSL) is emerging as a promising research topic where task-specific graph structure and GNN parameters are jointly learned in an end-to-end unified framework. Despite their great progress, existing approaches mostly focus on the design of similarity metrics or graph construction, but directly default to adopting downstream objectives as supervision, which lacks deep insight into the power of supervision signals. More importantly, these approaches struggle to explain how GSL helps GNNs, and when and why this help fails. In this article, we conduct a systematic experimental evaluation to reveal that GSL and GNNs enjoy consistent optimization goals in terms of improving the graph homophily. Furthermore, we demonstrate theoretically and experimentally that task-specific downstream supervision may be insufficient to support the learning of both graph structure and GNN parameters, especially when the labeled data are extremely limited. Therefore, as a complement to downstream supervision, we propose homophily-enhanced self-supervision for GSL (HES-GSL), a method that provides more supervision for learning an underlying graph structure. A comprehensive experimental study demonstrates that HES-GSL scales well to various datasets and outperforms other leading methods. Our code will be available in https://github.com/LirongWu/Homophily-Enhanced-Self-supervision.

Characteristic mango price forecasting using combined deep-learning optimization model.

Accurate product price forecasting is helpful for scientific decision-making and precise industrial planning. As a characteristic fruit that drives regional development, mango price prediction is of great significance to several economies. However, owing to the strong volatility of mango prices, forecasting is vulnerable to uncertainties and is very challenging. In this study, a deep-learning combination forecasting model based on a back-propagation (BP) long short-term memory (LSTM) neural network is proposed. Using daily mango price data from a large fruit wholesale trading center in China from January 2nd, 2014, to April 18th, 2022, mango price changes are learned and predicted to support the fruit industry. The results show that the root mean-square error, mean absolute percentage error, and the R2 determination coefficient of the BP-LSTM combination model are 0.0175, 0.14%, and 0.9998, respectively. The prediction results of the combined model are better than those of the separate BP and LSTM models. Furthermore, it best fits the actual price profile and has better generalizability.

Beyond Homophily and Homogeneity Assumption: Relation-Based Frequency Adaptive Graph Neural Networks.

Graph neural networks (GNNs) have been playing important roles in various graph-related tasks. However, most existing GNNs are based on the assumption of homophily, so they cannot be directly generalized to heterophily settings where connected nodes may have different features and class labels. Moreover, real-world graphs often arise from highly entangled latent factors, but the existing GNNs tend to ignore this and simply denote the heterogeneous relations between nodes as binary-valued homogeneous edges. In this article, we propose a novel relation-based frequency adaptive GNN (RFA-GNN) to handle both heterophily and heterogeneity in a unified framework. RFA-GNN first decomposes an input graph into multiple relation graphs, each representing a latent relation. More importantly, we provide detailed theoretical analysis from the perspective of spectral signal processing. Based on this, we propose a relation-based frequency adaptive mechanism that adaptively picks up signals of different frequencies in each corresponding relation space in the message-passing process. Extensive experiments on synthetic and real-world datasets show qualitatively and quantitatively that RFA-GNN yields truly encouraging results for both the heterophily and heterogeneity settings. Codes are publicly available at: https://github.com/LirongWu/RFA-GNN.

Effects of macromolecular organic acids on reducing inorganic phosphorus fixation in soil.

To improve the availability of inorganic phosphorus (P) in soil, we investigated the role of three macromolecular organic acids (MOAs), including fulvic acid (FA), polyaspartic acid (PA), and tannic acid (TA), in reducing the fixation of inorganic P fertilizer in the soil. AlPO4, FePO4, and Ca8H2(PO4)6·5H2O crystals were chosen as insoluble phosphate representatives in the soil to simulate the solubilization process of inorganic P by MOAs. The microstructural and physicochemical properties of AlPO4, FePO4, and Ca8H2(PO4)6·5H2O were determined by scanning electron microscopy (SEM), Fourier-transform infrared spectroscopy (FT-IR), and X-ray photoelectron spectroscopy (XPS) before and after treatment of MOAs. In addition, the amounts of leached P and fixed inorganic P in Inceptisols and Alfisols affected by MOAs combined with superphosphate (SP) fertilizer were determined by soil leaching experiments. The presence of the three MOAs significantly increased the concentration of leached P and reduced the contents of insoluble inorganic phosphate formed with iron, aluminum, and calcium fixed in the soil, in which PA combined with SP had the most significant effect. Furthermore, the less inorganic P fixation in the combination treatment of MOAs and SP resulted in a greater wheat yield and P uptake. Therefore, MOAs could be a synergistic material for increasing P fertilizer utilization.

A comprehensive investigation of a water and energy-based waste integrated system: Techno-eco-environmental-sustainability aspects.

In this work, a novel biomass gasifier combined energy system was offered for potable water, heating load, and power generation. The system included a gasifier, an S-CO2 cycle, a combustor, a domestic water heater, and thermal desalination unit. The plant was evaluated from various aspects, i.e., energetic, energetic, exergo-economic, sustainability, and environmental. To this aim, modeling of the suggested system was conducted by EES software; then, a parametric inquiry was carried out to detect the critical performance parameters, considering an environmental impact indicator. The results showed that the freshwater rate, Levelized CO2 emissions, total cost, and sustainability index of 21.19 kg s-1, 0.563 t.MWh-1, 13.13 $.GJ-1, and 1.53 were acquired, each. Moreover, the combustion chamber is a major fount in the irreversibility of the system. Besides, the energetic and exergetic efficiencies were computed at 89.51% and 40.87%. Overall, the offered water and energy-based waste system showed great functionality in terms of thermodynamic, economic, sustainability, and environmental standpoints by enhancing the gasifier temperature.

Microplastic pollution threats coastal resilience and sustainability in Xiamen City, China.

Microplastics have raised growing awareness due to their ubiquity and menaces to coastal resilience and sustainability. The abundance, distribution, and characteristics of microplastics in water and organisms in Xiamen were evaluated. Results showed that the average abundance of microplastics in the surface water of Xiamen Bay was 1.55 ± 1.94 items/m3. The dominant color, size, shape, and polymer type were white, 1.0-2.5 mm, and fragments and lines, and polyethylene and polypropylene, respectively. The average abundance of microplastics in the fish in Xiamen was 2.44 ± 1.56 items/g wet weight. They were dominated by fibers of blue polyethersulfone and polyethylene terephthalate, and sizes <2.5 mm. There was a negative correlation between the polymer type in fish and that in water, while a positive correlation between shapes of microplastics of both fish species. Results will aid in formulating management measures for preventing microplastic pollution in Xiamen, ultimately promoting coastal resilience and sustainability of coastal communities.

Microstructural Characterization of the As-Cast and Homogenized Al-Cu-Mg-Ag Alloy.

In this study, the as-cast microstructure and the evolution of the homogenized microstructure of large-scale industrialized Al-Cu-Mg-Ag heat-resistant aluminum alloy ingots were investigated by means of optical microscopy (OM), scanning electron microscopy (SEM), energy dispersive analysis (EDS), and differential scanning calorimetry (DSC). The results indicate that the dendritic segregation is evident in the ingot along the radial direction, and the grain boundaries are decorated with lots of net-shaped continuous eutectic structures. With the homogenization time extension and the homogenization temperature increase, the eutectic phases (i.e., the primary Al2Cu phase, the Al2CuMg phase, and the AlCuMgAg quaternary phase) at the grain boundaries gradually dissolve back into the matrix. Meanwhile, most of the dendritic grain boundaries gradually become sparse and thinner. Finally, it is found that the optimal homogenization regime of the Al-Cu-Mg-Ag alloy is 420 °C/5 h+480 °C/8 h+515 °C/24 h.

Transboundary microplastic pollution in Xiamen Bay and adjacent Jiulong River estuary after the outbreak of COVID-19.

Land-based transport from nearshore areas is a key pathway of microplastic (MP) pollution in the oceans. Therefore, transport, fate, and intervention on MPs necessitate an investigation of MP contamination in coastal regions. Here, MP pollution in the surface waters of Xiamen Bay and Jiulong River estuary was evaluated in 2021 after the outbreak of COVID-19. The abundance of MPs in Xiamen Bay ranged from 0.20 to 5.79 items m-3 with an average of 1.03 items m-3, whereas that in the Jiulong River estuary spanned from 0.55 to 2.11 items m-3 with a mean of 1.30 items m-3. A yearly decreasing trend in the abundance of MPs in surface waters in both regions was observed. The particle sizes of MPs were concentrated in the range of 2.50-5.00 mm, and the colors were mainly white, transparent, and green. The micro-Raman spectroscopic results showed that MP polymer types were predominantly polyethylene, polypropylene, and polystyrene. A lower abundance of MPs in Xiamen Bay with no obvious pattern was observed, while that in the Jiulong River estuary showed a wavelike distribution from upstream to downstream. Ecological risk assessment of MP pollution in surface waters of two regions was performed using the pollution load index (PLI), giving the risk level in descending order: wastewater discharge area > aquaculture area > sloughs > estuary mouth > estuarine rivers > shipping lane. The average risk level of Xiamen Bay (I) was lower than that in Jiulong River estuary (II). The MP pollution in the Jiulong River estuary appeared heavier than that in Xiamen Bay, which may be due to the combined effects of COVID-19 and marine governance. This study provided insights into the prevention and management of MP pollution in nearshore semi-enclosed bays.

Deep Clustering and Visualization for End-to-End High-Dimensional Data Analysis.

High-dimensional data analysis for exploration and discovery includes two fundamental tasks: deep clustering and data visualization. When these two associated tasks are done separately, as is often the case thus far, disagreements can occur among the tasks in terms of geometry preservation. Namely, the clustering process is often accompanied by the corruption of the geometric structure, whereas visualization aims to preserve the data geometry for better interpretation. Therefore, how to achieve deep clustering and data visualization in an end-to-end unified framework is an important but challenging problem. In this article, we propose a novel neural network-based method, called deep clustering and visualization (DCV), to accomplish the two associated tasks end-to-end to resolve their disagreements. The DCV framework consists of two nonlinear dimensionality reduction (NLDR) transformations: 1) one from the input data space to latent feature space for clustering and 2) the other from the latent feature space to the final 2-D space for visualization. Importantly, the first NLDR transformation is mainly optimized by one Clustering Loss, allowing arbitrary corruption of the geometric structure for better clustering, while the second NLDR transformation is optimized by one Geometry-Preserving Loss to recover the corrupted geometry for better visualization. Extensive comparative results show that the DCV framework outperforms other leading clustering-visualization algorithms in terms of both quantitative evaluation metrics and qualitative visualization.

Synthesis of Biodegradable Polyester-Polyether with Enhanced Hydrophilicity, Thermal Stability, Toughness, and Degradation Rate.

Novel poly(butylene succinate-butylene furandicarboxylate/polyethylene glycol succinate) (PBSF-PEG) was synthesized using two-step transesterification and polycondensation in the melt. There are characterized by intrinsic viscosity, GPC, 1H NMR, DSC, TGA, tensile, water absorption tests, and water degradation at different pH. GPC analysis showed that PBSF-PEG had high molecular weight with average molecular weight (Mw) up to 13.68 × 104 g/mol. Tensile tests showed that these polymers possessed good mechanical properties with a tensile strength as high as 30 MPa and elongation at break reaching 1500%. It should be noted that the increase of PEG units improved the toughness of the polyester material. In addition, the introduction of PEG promoted the water degradation properties of PBSF, and the copolymer showed a significantly faster water degradation rate when the PEG unit content was 20%. This suggests that the amount of PEG introduced could be applied to regulate the water degradation rate of the copolymers. Hence, these new polymers have great potential for application as environmentally friendly and sustainable plastic packaging materials.

Nontoxic Tb3+-induced hyaluronic nano-poached egg aggregates for colorimetric and luminescent detection of Fe3+ ions.

This study demonstrates that a luminescent Tb3+ complex with green emission can be complexed with hyaluronic (hya) to form nanoparticles. The structure of complexation is composed of a Tb(acac)2phen core with a hya surface, similar to those of the nano-poached eggs. What makes the structure unique is that Tb(acac)2phen and hya are connected by chemical bonds. To confirm their utility, we illustrate that the luminescence is rapidly and selectively quenched in the presence of Fe3+. Initial cytotoxicity experiments with human liver carcinoma cells show that the luminescent lanthanide complexes are cytotoxic, however, complexing lanthanides to hya renders them cytocompatible. The new complex integrates the advantages of superior lanthanide luminescence, the unique shape of nano-poached eggs, compatibility with aqueous systems, and cytocompatibility. Tb3+-induced hyaluronic nano-poached eggs (THNE) can, therefore, be used for Fe3+ detection in aqueous systems.

EP-ADTA: Edge Prediction-Based Adaptive Data Transfer Algorithm for Underwater Wireless Sensor Networks (UWSNs).

The underwater wireless sensor network is an important component of the underwater three-dimensional monitoring system. Due to the high bit error rate, high delay, low bandwidth, limited energy, and high dynamic of underwater networks, it is very difficult to realize efficient and reliable data transmission. Therefore, this paper posits that it is not enough to design the routing algorithm only from the perspective of the transmission environment; the comprehensive design of the data transmission algorithm should also be combined with the application. An edge prediction-based adaptive data transmission algorithm (EP-ADTA) is proposed that can dynamically adapt to the needs of underwater monitoring applications and the changes in the transmission environment. EP-ADTA uses the end-edge-cloud architecture to define the underwater wireless sensor networks. The algorithm uses communication nodes as the agents, realizes the monitoring data prediction and compression according to the edge prediction, dynamically selects the transmission route, and controls the data transmission accuracy based on reinforcement learning. The simulation results show that EP-ADTA can meet the accuracy requirements of underwater monitoring applications, dynamically adapt to the changes in the transmission environment, and ensure efficient and reliable data transmission in underwater wireless sensor networks.

Development of a machine learning model to predict early recurrence for hepatocellular carcinoma after curative resection.

Background: Early recurrence is common for hepatocellular carcinoma (HCC) after surgical resection, being the leading cause of death. Traditionally, the COX proportional hazard (CPH) models based on linearity assumption have been used to predict early recurrence, but predictive performance is limited. Machine learning models offer a novel methodology and have several advantages over CPH models. Hence, the purpose of this study was to compare random survival forests (RSF) model with CPH models in prediction of early recurrence for HCC patients after curative resection. Methods: A total of 4,758 patients undergoing curative resection from two medical centers were included. Fifteen features including age, gender, etiology, platelet count, albumin, total bilirubin, AFP, tumor size, tumor number, microvascular invasion, macrovascular invasion, Edmondson-Steiner grade, tumor capsular, satellite nodules and liver cirrhosis were used to construct the RSF model in training cohort. Discrimination, calibration, clinical usefulness and overall performance were assessed and compared with other models. Results: Five hundred survival trees were used to generate the RFS model. The five highest Variable Importance (VIMP) were tumor size, macrovascular invasion, microvascular invasion, tumor number and AFP. In training, internal and external validation cohort, the C-index of RSF model were 0.725 [standard errors (SE) =0.005], 0.762 (SE =0.011) and 0.747 (SE =0.016), respectively; the Gönen & Heller's K of RSF model were 0.684 (SE =0.005), 0.711 (SE =0.008) and 0.697 (SE =0.014), respectively; the time-dependent AUC (2 years) of RSF model were 0.818 (SE =0.008), 0.823 (SE =0.014) and 0.785 (SE =0.025), respectively. The RSF model outperformed early recurrence after surgery for liver tumor (ERASL) model, Korean model, American Joint Committee on Cancer tumor-node-metastasis (AJCC TNM) stage, Barcelona Clinic Liver Cancer (BCLC) stage and Chinese stage. The RSF model is capable of stratifying patients into three different risk groups (low-risk, intermediate-risk, high-risk groups) in the training and two validation cohorts (all P<0.0001). A web-based prediction tool was built to facilitate clinical application (https://recurrenceprediction.shinyapps.io/surgery_predict/). Conclusions: The RSF model is a reliable tool to predict early recurrence for patients with HCC after curative resection because it exhibited superior performance compared with other models. This novel model will be helpful to guide postoperative follow-up and adjuvant therapy.