Search

Based on the data of A-share listed companies in Shanghai and Shenzhen from 2012 to 2021, this paper tests the impact mechanism of ESG performance on the financial performance of environmentally sensitive enterprises under the mesomeric effect of green technology innovation by establishing a panel regression model. The research results indicate that there is a significant negative correlation between ESG performance and financial performance of environmentally sensitive enterprises, meaning that companies with better ESG performance tend to have lower financial performance. Green technological innovation plays a part of the mesomeric effect between the two. The research results provide important ideas and methods for environmentally sensitive enterprises to promote their financial performance in achieving high-quality economic development, and provide a theoretical basis for promoting the development of ESG and green technology innovation in China.

In previous taxonomic studies, identification of Atlanticus species was mainly based on male morphological features, such as cercus, tenth abdominal tergum and subgenital plate. In this study, we systematically revise eight Atlanticus species (61 individuals) in East China. We sequence COI genes of all samples for identification, and compare morphologies and male calling songs. The classification based on male cerci, calling songs, and COI genes is consistent. In contrast, the morphologies of male tenth abdominal tergum and subgenital plate varied among species. In addition, we regard A. jiangyei Liu, Wang & Cheng, 2016 as a junior synonym of A. pieli Tinkham, 1941, and A. helleri Liu, Wang & Cheng, 2016 as a junior synonym of A. kiangsu Ramme, 1939.

Cell-free DNA (cfDNA) analysis holds promise for early detection of lung cancer and benefits patients with higher survival. However, the detection sensitivity of previous cfDNA-based studies was still low to suffice the clinical use, especially for early-stage tumors.Establish an accurate and affordable approach for early-stage lung cancer detection by integrating cfDNA fragmentomics and machine learning models.This study included 350 non-cancer and 432 cancer participants. The participants' plasma cfDNA samples were profiled by whole-genome sequencing. Multiple cfDNA features and machine learning models were compared in the training cohort to achieve an optimal model. Model performance was evaluated in three validation cohorts.A stacked ensemble model integrating five cfDNA features and five machine learning algorithms constructed in the training cohort (cancer: 113, healthy: 113) outperformed all the models built on individual feature-algorithm combinations. This integrated model yielded superior sensitivities of 91.4% at 95.7% specificity for Cohort Validation I [Area Under the Curve (AUC): 0.984], 84.7% at 98.6% specificity for Validation II (AUC: 0.987), and 92.5% at 94.2% specificity for Additional Validation (AUC: 0.974), respectively. The model's high performance remained consistent when sequencing depth was down to 0.5× (AUC: 0.966-0.971). Furthermore, our model is sensitive to identifying early pathological features (83.2% sensitivity for stage I, 85.0% sensitivity for1cm tumor at the 0.66 cutoff).We have established a stacked ensemble model using cfDNA fragmentomics features and achieved superior sensitivity for detecting early-stage lung cancer, which could promote early diagnosis and benefit more patients. This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives License 4.0 (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Clusterwise linear regression (CLR) is a well-known technique for approximating a data using more than one linear function. It is based on the combination of clustering and multiple linear regression methods. This article provides a comprehensive survey and comparative assessments of CLR including model formulations, description of algorithms, and their performance on small to large-scale synthetic and real-world datasets. Some applications of the CLR algorithms and possible future research directions are also discussed.

Background. Pulmonary artery hypertension (PAH) is a common complication of congenital heart disease (CHD) and is associated with worse outcomes and increased mortality. The Doppler echocardiography (DE) is a commonly used imaging tool for both diagnosis and follow-up examination of PAH. Here is to evaluate the diagnostic performance of DE combined with NTproBNP/BNP as screening strategy in PAH patients with CHD. Methods. A retrospective study in 64 patients with CHD has been carried out to compare estimate pulmonary artery systolic pressure (PASP) measured with DE to that measured with right heart catheterization (RHC). The Pearson correlation analyses were used to calculate the correlation coefficients between RHC and DE. The Bland-Altman analyses were carried out to assess the agreement between the two methods. ROC analyses were used to evaluate the diagnostic performance of DE, NTproBNP/BNP, and DE combined with NTproBNP/BNP. Results. Our data have demonstrated that a mild correlation ( r = 0.4401 , P < 0.01 ) was observed between PASP ( 78.1 ± 29.0 mmHg) measured during RHC and PASP ( 74.9 ± 19.7 mmHg) as estimated using DE. The Bland-Altman analysis demonstrated that the bias for DE PASP estimates was 3.2 mmHg with 95% limits of agreement ranging from -49.53 to 55.90 mmHg. The results of DE showed an AUC of 0.848 (95% CI = 0.666 -1; P < 0.001 ), the sensitivity of which was 98.3% and the specificity was 77.8%. The AUC of NTproBNP/BNP for the identification of PAH was 0.804 (95% CI = 0.651 -0.956; P < 0.001 ), the sensitivity of which was 81.4% and the specificity was 87.5%. The AUC of DE combined with NTproBNP/BNP was 0.857 (95% CI = 0.676 -1; P < 0.001 ), of which sensitivity was 100% and specificity was 77.8%. The positive predictive value (PPV) and negative predictive value (NPV) were 96.6% and 100%, respectively. Conclusions. Our study shows that the Doppler echocardiography combined with NTproBNP/BNP has better diagnostic performance in pulmonary artery hypertension associated with congenital heart disease, especially when DE negative screening in PAH patients.

Two novel Cd(ii)-based metal–organic frameworks are constructed from an imidazole-containing ligand and SiF62− anions, and both of them exhibit a fluorescence sensing capacity for arginine via a turn-on effect.

In order to clarify the temporal and spatial changes in the internal photothermal environment in an open-field agrivoltaic system (OAVS), this paper took the OAVS in eastern China as the research object and divided the internal area into the southern area, middle area and northern area, according to the spatial structure. Further, a photothermal environment test was conducted in the above three areas in the summer and winter. The results showed that the summer average daylight rate (Rm-avg) in the middle area was 66.6%, while the Rm-avg in the other two areas was about 20%, with no significant difference. In the winter, the light environment in the southern area was slightly better, and the Rm-avg in the above three areas was 26.4%, 24.7% and 19.7%, respectively. On the whole, the relationship between the thermal environmental factors and the solar radiation intensity was consistent. Further, a 3D model of an OAVS was established using Autodesk Ecotect Analysis 2011, and the internal light environment was simulated. Compared with the measured values, the relative error was less than 10%, which verified the reliability of the OAVS model. Then, the model was used to reveal the temporal and spatial changes in the light environment of the OAVS. The simulation results showed that the daylighting rate in the summer from the ground to the height of the fig canopy inside the system was 20.7% to 61.5%. In the winter, the daylighting rate from the ground to the height of the fig canopy inside the system was 17.7% to 36.4%. The effectiveness of the OAVS in reducing the level of solar radiation intensity depended on the time of day and the angle of the sun. At the spatial scale, due to the strong consumption of light by photovoltaic panels, there was a strong horizontal and vertical light environment gradient inside the system. In conclusion, the photothermal environment research of an OAVS based on Autodesk Ecotect Analysis 2011 can not only provide a basis for agricultural production and structural design such as span, height and the laying density of PV panels, but also expand its application to regions with different latitudes and longitudes and specific climates.

We aimed to investigate levels of cytokines/chemokines and immune checkpoint molecules in patients with anti-leucine-rich glioma-inactivated 1 (LGI1) encephalitis.The study recruited 12 patients with anti-LGI1 encephalitis and six non-inflammatory controls from the Qilu Hospital of Shandong University treated between January 2019 and December 2020. Serum levels of 30 cytokines/chemokines and 10 checkpoint molecules were measured in participants of both the groups.In contrast to those in the control group, 24 cytokines/chemokines and 5 immune checkpoint molecules were differentially expressed in patients with anti-LGI1 encephalitis, with 14 cytokines being upregulated and 10 being downregulated. There were 1033 enriched biological processes and 61 enriched Kyoto Encyclopedia of Genes and Genomes signaling pathways.A wide range of cytokines/chemokines and immune checkpoint molecules are implicated in immune regulation in anti-LGI1 encephalitis, indicating that they may serve as important targets in the development and treatment of the disease.

The aim was to evaluate the potential of retinal nerve fiber layer thickness (RNFLT) measured with optical coherence tomography in predicting disease progression in relapsing-remitting multiple sclerosis (RRMS).Analyses were conducted post hoc of this 24-month, phase III, double-blind study, in which RRMS patients were randomized (1:1:1) to once daily oral fingolimod 0.5 mg, 1.25 mg or placebo. The key outcomes were the association between baseline RNFLT and baseline clinical characteristics and clinical/imaging outcomes up to 24 months. Change of RNFLT with fingolimod versus placebo within 24 months and time to retinal nerve fiber layer (RNFL) thinning were evaluated.Altogether 885 patients were included. At baseline, lower RNFLT was correlated with higher Expanded Disability Status Scale score (r = -1.085, p = 0.018), lower brain volume (r = 0.025, p = 0.006) and deep gray matter volume (r = 0.731, p 0.0001), worse visual acuity (r = -19.846, p 0.0001) and longer duration since diagnosis (r = -0.258, p = 0.018). At month 12, low baseline RNFLT (86 μm) versus high baseline RNFLT (≥99 μm) was associated with a greater brain volume loss (percentage change -0.605% vs. -0.315%, p = 0.035) in patients without optic neuritis history. At month 24, low baseline RNFLT versus high baseline RNFLT was associated with a higher number of new or newly enlarged T2 lesions (mean number 4.0 vs. 2.8, p = 0.014) and a higher risk of subsequent RNFL thinning (hazard ratio 2.55; 95% confidence interval 1.84-3.53; p 0.001). The atrophy of the RNFL in the inferior quadrant was alleviated with fingolimod 0.5 mg versus placebo at month 24 (Δ(least squares mean) = 1.8, p = 0.047).Retinal nerve fiber layer thickness could predict disease progression in RRMS.Clinicaltrials.gov identifier: NCT00355134, https://clinicaltrials.gov/ct2/show/NCT00355134.

This study aims to improve regional agricultural production efficiency and promote sustainable agricultural development by presenting a data-driven evaluation method for regional agricultural production efficiency. Based on data collection and processing of regional agricultural input-output factors in Anhui Province, China, from 2014 to 2019, a data envelopment analysis Malmquist model is constructed for data modeling. Static analysis of regional agricultural production efficiency and production redundancy is conducted, and the dynamic change of regional agricultural production efficiency is measured. The results show that technical efficiency is the core driving factor for improving regional agricultural production efficiency. The findings indicate significant policy implications for improving agricultural production efficiency from the perspective of regional agricultural high-quality development. This study provides theoretical and methodological support for the sustainable development of regional agriculture.

Kinase domain duplications (KDDs) have recently been recognized as oncogenic mutations and possible association with drug resistance in cancers.Here, targeted sequencing was performed with the tumor tissue and/or plasma from 65 cancer patients with KDDs.Intact KDDs were identified in approximately 0.1% of the total population across multiple cancer types. EGFR KDD was first identified in colorectal cancer and breast cancer, whereas FGFR2 KDD was first identified in gastric cancer. Tumors with EGFR KDD displayed lower concurrent TP53 gene alterations (p = 0.03) and slightly higher chromosome instability (p = 0.27) compared to tumors with non-EGFR-KDDs. Immune pathway analysis further revealed the enrichment of the cytokine receptors pathway (93%) in the KDD carriers. Hyperprogression-related gene mutations were identified in four cases.Collectively, our data revealed the genomic features of KDD alterations in a multi-cancer cohort, providing more information for the potential treatment application in the KDD carriers.

With rare genetic variations having been increasingly recognized at a preclinical stage, a variety of early-phase clinical trials have been launched. Due to the low incidence rate of these variations, although the sample size of trials are small, it still needs a large number of patients for screening. With the advent of next-generation sequencing (NGS), multiple genetic variations can be detected simultaneously. Multiple biomarkers and agents can be evaluated using umbrella clinical trials, which rapidly and effectively screen and enroll patients for parallel sub-studies using NGS.We designed an open-label, multi-center, phase II clinical trial CTONG1702. This is an adaptive umbrella trial that will evaluate the efficacy and safety of several biomarker-driven agents, including tyrosine kinase inhibitors (TKIs) and a PD-1 inhibitor, in stage IIIB to IV patients (eighth AJCC) with non-small-cell lung cancer (NSCLC) patients. Patients will be enrolled in parallel sub-studies based on the results of NGS and PD-L1 IHC analysis. Patients who are not eligible for CTONG1702 will be enrolled in the observational real-world study CTONG1705. This study aims to develop a large-scale genomic database and explore the relationship between genetic variations in NSCLC patients and clinical outcomes.The adaptive umbrella trial will evaluate multi-targets and multi-drugs in advanced NSCLC patients (CTONG1702). In addition, the simultaneously initiated real-world study will provide additional data for clinical practice (CTONG1705).

Background A "Watch and Wait” (W&W) approach has become an alternative to surgery for locally advanced rectal cancer (LARC) after neoadjuvant chemoradiotherapy (nCRT). Precise prediction of pathological complete response (pCR) will improve patient selection for W&W. We investigated the utility of cell-free DNA (cfDNA) fragmentomics in predicting pCR. Methods We recruited 119 LARC patients and evaluated nCRT response by pCR status and pathological or MRI tumor regression grade (mrTRG). Plasma samples before, during, and after nCRT were applied to deep targeted-panel sequencing, with 103 patients having complete samples. cfDNA fragment and 5′-end motif profiles were used to construct elastic-net logistic regression models to predict non-pCR. Predictive performance was measured by area under the receiver operator characteristic curve (AUC), sensitivity, and specificity. Results In the training cohort, the model based on 5′-end motif profile plus mrTRG achieved the highest cross-validation AUC (0.92, 95% CI, 0.91–0.93). The AUC in a testing cohort was 0.96 (95% CI, 0.90–1.00). The models based on 5′-end motif profile alone or in combination with mrTRG both maintained good predictive ability for patients without detectable circulating tumor DNA (AUC 0.94, 95% CI, 0.93–0.95; AUC 0.95, 95% CI, 0.94–0.96). In an external validation cohort, the model trained with a local 5′-end motif profile obtained an AUC of 0.878 (95% CI, 0.801–0.956) in discriminating colorectal cancer from healthy subjects. Conclusions The combination of a 5′-end motif profile with mrTRG has the potential to predict the response to nCRT, and therefore may improve the patient selection for a W&W approach.

In recent years, the term “big data” has attracted the attention of many scholars and business managers, and the emergence of massive data has ushered in a major transformation of the times. Service quality and customer repeat purchase intention are two of the hot issues in the field of service research. Clarifying the mechanism of the two can help enterprises establish long-term customer relations with more customers through the improvement of service quality, enhance their competitive strength, and improve enterprise performance. According to the actual consumer purchase data provided, the existing data sets are visually analyzed and processed to find the consumer purchase rules, build features and build a reasonable consumer purchase prediction experimental data set. Based on the obtained experimental data set, a single prediction model of consumer purchase is designed. According to the analysis of the instantiation results of the single prediction model, the single model consumed in the fusion model of consumer purchase prediction is determined. In order to demonstrate the viewpoints put forward, the research model is established, the research hypothesis is put forward, the questionnaire is designed by referring to the literature and field investigation, and the results of the questionnaire are empirically analyzed. Using the methods of descriptive statistical analysis, factor analysis, reliability and validity test, and regression analysis, the hypothesis that service quality has a direct impact on customer repeat purchase intention is verified.

Integrating metal-organic framework (MOF)-covalent organic framework (COF) allows versatile engineering of hybrid materials with properties superior to pristine components, especially COFs suffered from aggregation-caused quenching (ACQ), unlocking more possibilities to improve the luminescence of COFs. In this work, we prepared various MOF@COF composites with different COF layer thicknesses, in which stable UiO-66-NH

Adriamycin (ADR) is an efficient and common broad-spectrum anticancer drug. However, the cumulative and dose-dependent toxicity induced by ADR severely limits its application in the clinic. Previous studies found that psoralidin (PSO) exhibits remarkable therapeutic effects against multiple cancers.The aim of this study was to determine if PSO has beneficial effects on ADR-induced cardiotoxicity and to investigate the underlying mechanisms.ADR-induced cardiotoxicity models were established in BALB/c mice and HL-1 cardiomyocytes. A series of experimental methods were used to evaluate the effects of PSO on cardiac function indicators, blood biochemical parameters, histopathology, oxidative stress, apoptosis, mitochondrial function, fibrosis, and SIRT1/PPARγ signaling.PSO significantly improved cardiac function indicators, blood biochemical parameters, and mitochondrial function and reduced the degree of myocardial fibrosis, oxidative stress, and apoptosis in ADR-injured mice. PSO significantly increased cell viability, inhibited the release of LDH, reduced oxidative stress and apoptosis, and improved mitochondrial function in ADR-injured HL-1 cells. Moreover, we also demonstrated there was cross-talk between SIRT1 and PPARγ, as shown by SIRT1 siRNA significantly decreasing the expression of PPARγ and GW9662 (a PPARγ antagonist), which remarkably reduced the expression of SIRT1.In summary, this study proved for the first time the beneficial effect of PSO on ADR-induced cardiotoxicity through activation of the SIRT1/PPARγ signaling pathway. Therefore, these findings may favor PSO as a potential cardioprotective drug candidate to alleviate ADR-induced cardiotoxicity in the clinic and improve the application of ADR in oncotherapy.

Based on the quality-controlled observational spectral width data of the Beijing mesosphere–stratosphere–troposphere (MST) radar in the altitudinal range of 3–19.8 km from 2012 to 2014, this paper analyses the relationship between the proportion of negative turbulent kinetic energy (N-TKE) and the horizontal wind speed and the vertical shear of horizontal wind domain and gives the distributional characteristics of atmospheric turbulence parameters obtained by using different calculation models. Three calculation models of the spectral width method were used in this study – namely the H model (Hocking, 1985), N-2D model (Nastrom, 1997) and D–H model (Dehghan and Hocking, 2011). The results showed that the proportion of N-TKE in the H model, N-2D model and D–H model increases with the horizontal wind speed u and/or the vertical shear of horizontal wind speed ∂u∂z, and the maximum values are 60 %, 45 % and 35 %, respectively. When the∂u∂z is greater than 0.006 s−1, the N-TKE of the H model increases sharply with ∂u∂z; the increase rate is about 20%0.002s-1. For these three models, the results are similar except that the vertical shear of the horizontal wind speed is greater than 0.006 s−1. When ∂u∂z>0.006 s−1, the proportion of N-TKE in the N-2D and H models increases with ∂u∂z, while the proportion in the D–H model is less than 10 % and has slight variation. However, it is still necessary to consider the applicability of the N-2D model and D–H model in some weather processes with strong winds. The distributional characteristics with height of the turbulent kinetic energy dissipation rate ε and the vertical eddy diffusion coefficient Kz derived by the three models are consistent with previous studies. Still, there are differences in the values of turbulence parameters. Also, the range resolution of the radar has little effect on the differences in the range of turbulence parameters' values. The median values of ε in the H model, N-2D model and D–H model is 10−3.2–10−2.7, 10−3.0–10−2.6 and 10−3.3–10−2.8 m2 s−3, respectively. The median values of Kz in these three models are 100.3–100.7, 100.4–100.7 and 100.1–100.5 m2 s−1.

With the increasing severity of the global energy crisis and environmental pollution, there is an urgent need to change the economic development model driven by certain factors and the investment scale and pursue science- and technology-driven innovative development. This study aims to improve the efficiency of scientific and technological innovation and promote the high-quality development of regional industrial enterprises. It constructs a data-driven DEA-Malmquist evaluation model to evaluate and optimize regional industrial enterprises’ scientific and technological innovation efficiency. First, we collect the panel data of regional industrial enterprises’ scientific and technological innovation input-output indexes. Second, we use the Pearson correlation coefficient method to identify and construct the evaluation index system of regional industrial enterprises’ scientific and technological innovation efficiency. Third, we build a DEA-Malmquist evaluation model to quantitatively evaluate regional industrial enterprises’ scientific and technological innovation efficiency from static and dynamic aspects. Finally, we verify the feasibility and effectiveness of the method using statistical data on scientific and technological innovation and development of Anhui industrial enterprises from 2011 to 2019 and put forth targeted countermeasures and suggestions. This study provides theoretical and methodological support for the sustainable development of industrial enterprises.

Background Tumour immune microenvironment heterogeneity is prevalent in numerous cancers and can negatively impact immunotherapy response. Immune heterogeneity and evolution in gastroesophageal adenocarcinoma (GEA) have not been studied in the past. Methods Together with a multi-region sampling of normal, primary and metastatic tissues, we performed whole exome sequencing, TCR sequencing as well as immune cell infiltration estimation through deconvolution of gene expression signals. Results We discovered high TCR repertoire and immune cell infiltration heterogeneity among metastatic sites, while they were homogeneous among primary and normal samples. Metastatic sites shared high levels of abundant TCR clonotypes with blood, indicating immune surveillance via blood. Metastatic sites also had low levels of tumour-eliminating immune cells and were undergoing heavy immunomodulation compared to normal and primary tumour tissues. There was co-evolution of neo-antigen and TCR repertoire, but only in patients with late diverging mutational evolution. Co-evolution of TCR repertoire and immune cell infiltration was seen in all except one patient. Conclusions Our findings revealed immune heterogeneity and co-evolution in GEA, which may inform immunotherapy decision-making.

As industry is the world's leading carbon emitter, promoting industrial carbon reduction is of key significance to carbon peak and carbon neutrality. Using a data-driven method, based on the collection and processing of relevant data from statistical yearbooks and others, we analyze the efficiency and amount of carbon emission of each industrial sector after processing multi-dimensional data by the improved IPCC EF method of calculating carbon emissions. In addition, we adopt the LMDI decomposition method for data modeling to measure the contribution of energy efficiency, industrial structure, GDP per capita, and population size to carbon emission changes, to identify targets for industrial carbon reduction, and to propose a targeted optimization path for carbon emission. We show how the method is implemented by taking the statistics of Anhui Province from 2010 to 2019 as an example and advises on an optimization path for carbon emission in Anhui Province. This study is of both theoretical and practical significance as it provides theoretical and methodological support for the low-carbon development of the regional industry, and provides a reference for other countries and regions to explore the path of low-carbon and environment-friendly green transformation and upgrading.

BackgroundEsophageal cancer (EC) is an aggressive neoplasm of the gastrointestinal tract that is usually treated with a combination of chemotherapy, radiotherapy (RT), and/or surgery, according to disease status. Despite the availability of multimodal therapeutic strategies, local recurrence is frequently observed. However, there is no standard treatment or promising therapeutic approach for local recurrence or metastatic esophageal carcinoma after the RT. This study tended to investigate the efficacy and safety of sintilimab maintenance after concurrent chemoradiotherapy (CCRT) for local/regional recurrent esophageal squamous carcinoma.MethodsThis study was a single-arm, phase Ib/II trial conducted in a single site in China. Patients previously radically treated (surgery or CCRT), histologically confirmed, local or regional recurrence esophageal squamous carcinoma, qualified for the study design, were treated with 25–28 times radiotherapy plus raltitrexed once every 3 weeks for up to two cycles. Patients who have not progressed after CCRT received sintilimab as maintenance once every 3 weeks up to 1 year. Primary endpoints were overall survival (OS) and safety. Secondary endpoints were progression-free survival (PFS), objective response rate (ORR), and duration of response (DOR).ResultsBetween September 2019 and March 2022, in a total of 36 enrolled patients, 34 pts completed CCRT. Three patients excluded due to violation of the exclusion criteria (1 pt) and consent withdrawal (2 pts). Finally, 33 pts were included in the final analysis, in which 3 pts had disease progression, and the remaining 30 entered maintenance therapy with sintilimab. The median follow-up time was 12.3 months. Median OS was 20.6 months (95%CI 10.5–NA) and the 1-year OS rate was 64%. Median PFS was 11.5 months (95%CI 5.29–21.3) and the 1-year PFS rate was 43.6%. The ORR was 63.6% (95%CI 44.6–77.8), including 2 cases of CR and 19 cases of PR. The DCR was 19.9%, the median DOR was 19.5 months, and the median TTR was 2.4 months. The rate of any grade TRAEs was 96.7%; ≥Grade 3 TRAE was 23.4%. The incidence of immune-related AE was 60%, most of which were grade 1–2, and only one case of thyroid-stimulating hormone increased was irAE with grade 3 or above.ConclusionSintilimab has shown promising clinical efficacy and a manageable safety profile as maintenance therapy after CCRT for local/regional recurrent esophageal squamous carcinoma. In addition, further confirmation from a large-scale real-world study is still needed.

Identifying the stable structures of gold (Au) clusters is a huge challenge in cluster science. In this work, we have searched the ground-state structures of neutral Aun (n = 16 – 25) clusters by an artificial neural network (ANN) potential trained to density functional theory (DFT) data. Compared with the DFT data, the root mean square error of binding energy predicted by the ANN potential is about 8.66 meV/atom. Applying the ANN potential to search the ground-state structures by comprehensive genetic algorithm, we have found several new candidates of Au18, Au22, and Au23, which are not reported before. Au18 is hollow cage structure, Au22 and Au23 are flat cage structures. From the electronic analysis, we elucidate the stability mechanism of the newly found structures that are associated with the electronic shell closure of superatomic orbitals. Besides, we also clarified how to clean a database to train an efficient ANN potential in details. Overall, this work proves that applying machine learning in the description of atomic interactions can accelerate the searching of ground state structures of clusters and help to find new candidates of stable cluster structures.

Supplementary Table S1. Patient demography Supplementary Table S2. Evaluating fragmentomics model performances and ctDNA mutation-based method.

Currently, ~30–55% of the non-small cell lung cancer (NSCLC) patients develop recurrence due to minimal residual disease (MRD) after receiving surgical resection of the tumor. This study aims to develop an ultra-sensitive and affordable fragmentomic assay for MRD detection in NSCLC patients. A total of 87 NSCLC patients, who received curative surgical resections (23 patients relapsed during follow-up), enrolled in this study. A total of 163 plasma samples, collected at 7 days and 6 months postsurgical, were used for both whole-genome sequencing (WGS) and targeted sequencing. WGS-based cell-free DNA (cfDNA) fragment profile was used to fit regularized cox regression models, and leave-one-out cross-validation was further used to evaluate models' performance. The models showed excellent performances in detecting patients with a high risk of recurrence. At 7 days postsurgical, the high-risk patients detected by our model showed an increased risk of 4.6 times, while the risk increased to 8.3 times at 6 months post-surgical. These fragmentomics determined higher risk compared to the targeted sequencing based circulating mutations both at 7 days and 6 months postsurgical. The overall sensitivity for detecting patients with recurrence reached 78.3% while using both fragmentomics and mutation results from 7 days and 6 months postsurgical, which increased from the 43.5% sensitivity by using only the circulating mutations. The fragmentomics showed great sensitivity in predicting patient recurrence compared to the traditional circulating mutation, especially after the surgery for early-stage NSCLC, therefore exhibiting great potential to guide adjuvant therapeutics.

Additional methods for Coxnet models