A mega study of antibiotics contamination in Eastern aquatic ecosystems of China: occurrence, interphase transfer processes, ecotoxicological risks, and source modeling.

Understanding the occurrence, sources, transfer mechanisms, fugacity, and ecotoxicological risks of antibiotics play a pivotal role in improving the sustainability and ecological health of freshwater ecosystems. Therefore, in order to determine the levels of antibiotics, water and sediment samples were collected from multiple Eastern freshwater ecosystems (EFEs) of China, including Luoma Lake (LML), Yuqiao Reservoir (YQR), Songhua Lake (SHL), Dahuofang Reservoir (DHR), and Xiaoxingkai Lake (XKL), and were analyzed using Ultra Performance Liquid Chromatography/Tandem Mass Spectrometry (UPLC-MS/MS). EFEs regions are particularly interesting due to higher urban density, industrialization, and diverse land use in China. The findings revealed that a collective total of 15 antibiotics categorized into four families, which included sulfonamides (SAs), fluoroquinolones (FQs), tetracyclines (TCs), and macrolides (MLs), exhibited high detection frequencies, indicating widespread antibiotic contamination. The pollution levels in the water phase were in the order of LML > DHR > XKL > SHL > YQR. The sum concentration of individual antibiotics for each water body ranged from not detected (ND) to 57.48 ng/L (LML), ND to 12.25 ng/L (YQR), ND to 57.7 ng/L (SHL), ND to 40.50 ng/L (DHR), and ND to 26.30 ng/L (XKL) in the water phase. Similarly, in the sediment phase, the sum concentration of individual antibiotics ranged from ND to 15.35 ng/g, ND to 198.75 ng/g, ND to 1233.34 ng/g, ND to 388.44 ng/g, and ND to 862.19 ng/g, for LML, YQR, SHL, DHR, and XKL, respectively. Interphase fugacity (ffsw) and partition coefficient (Kd) indicated dominant resuspension of antibiotics from sediment to water, causing secondary pollution in EFEs. Two groups of antibiotics, namely MLs (erythromycin, azithromycin, and roxithromycin) and FQs (ofloxacin and enrofloxacin), showed a medium-high level of adsorption tendency on sediment. Source modeling (PMF5.0) identified wastewater treatment plants, sewage, hospitals, aquaculture, and agriculture as the major antibiotic pollution sources in EFEs, contributing between 6% and 80% to different aquatic bodies. Finally, the ecological risk posed by antibiotics ranged from medium to high in EFEs. This study offers valuable insights into the levels, transfer mechanisms, and risks associated with antibiotics in EFEs, enabling the formulation of large-scale policies for pollution control.

Levels, Distributions, and Potential Risks of Hexachlorobutadiene from Two Tetrachloroethylene Factories in China.

A systematic investigation was conducted on the emission of hexachlorobutadiene (HCBD) from two tetrachloroethylene factories that used the acetylene method (F1) and the tetrachloride transformation method (F2). The levels of HCBD in the air for F1 were found to be in the range of 1.46-1170 µg/m3, whereas F2 had levels in the range of 1.96-5530 µg/m3. Similarly, the levels of HCBD in the soil for F1 were found to be in the range from 42.2 to 140 µg/kg, whereas F2 had levels in the range from 4.13 to 2180 µg/kg. Samples obtained from the air, soil, and sludge in the reaction area of the tetrachloroethylene factories in China showed high levels of HCBD. The F1 method unintentionally produced more HCBD than the F2 method during tetrachloroethylene production, leading to greater harm. The results of the risk assessment suggested the presence of harmful health effects on workers in the workplace. The investigation findings highlight the need for improved management systems to ensure the safe production of tetrachloroethylene.

ABT-MPNN: an atom-bond transformer-based message-passing neural network for molecular property prediction.

Graph convolutional neural networks (GCNs) have been repeatedly shown to have robust capacities for modeling graph data such as small molecules. Message-passing neural networks (MPNNs), a group of GCN variants that can learn and aggregate local information of molecules through iterative message-passing iterations, have exhibited advancements in molecular modeling and property prediction. Moreover, given the merits of Transformers in multiple artificial intelligence domains, it is desirable to combine the self-attention mechanism with MPNNs for better molecular representation. We propose an atom-bond transformer-based message-passing neural network (ABT-MPNN), to improve the molecular representation embedding process for molecular property predictions. By designing corresponding attention mechanisms in the message-passing and readout phases of the MPNN, our method provides a novel architecture that integrates molecular representations at the bond, atom and molecule levels in an end-to-end way. The experimental results across nine datasets show that the proposed ABT-MPNN outperforms or is comparable to the state-of-the-art baseline models in quantitative structure-property relationship tasks. We provide case examples of Mycobacterium tuberculosis growth inhibitors and demonstrate that our model's visualization modality of attention at the atomic level could be an insightful way to investigate molecular atoms or functional groups associated with desired biological properties. The new model provides an innovative way to investigate the effect of self-attention on chemical substructures and functional groups in molecular representation learning, which increases the interpretability of the traditional MPNN and can serve as a valuable way to investigate the mechanism of action of drugs.

A machine learning model trained on a high-throughput antibacterial screen increases the hit rate of drug discovery.

Screening for novel antibacterial compounds in small molecule libraries has a low success rate. We applied machine learning (ML)-based virtual screening for antibacterial activity and evaluated its predictive power by experimental validation. We first binarized 29,537 compounds according to their growth inhibitory activity (hit rate 0.87%) against the antibiotic-resistant bacterium Burkholderia cenocepacia and described their molecular features with a directed-message passing neural network (D-MPNN). Then, we used the data to train an ML model that achieved a receiver operating characteristic (ROC) score of 0.823 on the test set. Finally, we predicted antibacterial activity in virtual libraries corresponding to 1,614 compounds from the Food and Drug Administration (FDA)-approved list and 224,205 natural products. Hit rates of 26% and 12%, respectively, were obtained when we tested the top-ranked predicted compounds for growth inhibitory activity against B. cenocepacia, which represents at least a 14-fold increase from the previous hit rate. In addition, more than 51% of the predicted antibacterial natural compounds inhibited ESKAPE pathogens showing that predictions expand beyond the organism-specific dataset to a broad range of bacteria. Overall, the developed ML approach can be used for compound prioritization before screening, increasing the typical hit rate of drug discovery.

Multi-isotope tracing nitrate dynamics and sources during thermal stratification in a deep reservoir.

Excessive nitrate (NO3-) input to reservoirs is a global concern. However, the dynamics and sources of NO3- under thermal stratification in deep reservoirs were rarely explored. In this study, multi-stable isotopes (δ15N/δ18O-NO3-, δ15N-particulate nitrogen (PN), δ15N-dissolved total nitrogen (DTN), and δ2H/δ18O-H2O) and a Bayesian mixing model were applied to reveal the biogeochemical processes and sources of NO3- in a deep reservoir with obvious nitrogen pollution. The results showed that the reservoir was thermally stratified in July while vertically mixed in October. The distribution of δ2H-H2O suggested that riverine nitrogen migrated to the epilimnion and metalimnion during stratification in the reservoir. In the epilimnion and metalimnion, the significant reduction in NO3- concentration was related to the enhancement of assimilation by thermal stratification. Meanwhile, the positive linear correlations between δ18O-NO3- and δ18O-H2O suggested that in-reservoir nitrification occurred, with its depth confined above the hypolimnion. In the hypolimnion, denitrification processes were absent due to the aerobic environment. Overall, NO3- dynamics were mainly controlled by nitrogen inflow, in-reservoir nitrification, and assimilation during thermal stratification. The results of the Bayesian mixing model showed that manure and sewage, and soil nitrogen were the dominant NO3- sources of the reservoir. This study provides new insights and data to help manage and restore deep waters worldwide in tackling a similar situation of nitrogen contamination.

The Role of SETBP1 in Gastric Cancer: Friend or Foe.

Background: Gastric cancer (GC) remains a common disease with a poor prognosis worldwide. The SET binding protein 1 (SETBP1) has been implicated in the pathogenesis of several cancers and plays a dual role as an oncogene and a tumor suppressor gene. However, the role and underlying mechanism of SETBP1 in GC remain unclear. Materials and Methods: We used next-generation RNA sequencing (RNA-seq) data from The Cancer Genome Atlas (TCGA) to explore the correlation between SETBP1 expression and tumor progression. We then quantified SETBP1 expression in GC cells with real-time quantitative polymerase chain reactions (RT-qPCR). The chi-square test and logistic regression were used to assess the correlation between SETBP1 expression and clinicopathological features. Kaplan-Meier survival analysis and Cox proportional hazards regression model were used to assess the relationship between SETBP1 expression and survival. Finally, gene set enrichment analyses (GSEA) were used to examine GC-related signaling pathways in low and high SETBP1 expressing samples. Results: We found SETBP1 expression levels in GC tissues to be significantly lower than in adjacent non-tumor tissues in the TCGA database. In addition, SETBP1 expression differed significantly between groups classified by tumor differentiation. Furthermore, SETBP1 expression in diffuse-type GC was significantly higher than in intestinal-type GC. However, it did not differ significantly across pathological- or T-stage groups. RT-qPCR and comprehensive meta-analysis showed that SETBP1 expression is downregulated in GC cells and tissues. Interestingly, SETBP1 expression in poorly- or un-differentiated GC cells was higher than in well-differentiated GC cells. Moreover, the chi-square test and logistic regression analyses showed that SETBP1 expression correlates significantly with tumor differentiation. Kaplan-Meier curves indicated that patients with relatively high SETBP1 expression had a poor prognosis. Multivariate analyses indicated that SETBP1 expression might be an important predictor of poor overall survival in GC patients. GSEA indicated that 20 signaling pathways were significantly enriched in samples with high and low SETBP1 expression. Conclusion: SETBP1 may play a dual role in GC progression.

A mathematical-adapted model to analyze the characteristics for the mortality of COVID-19.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) emerged in Wuhan, China, has led to the rapid development of Coronavirus disease 2019 (COVID-19) pandemic. COVID-19 represents a fatal disease with a great global public health importance. This study aims to develop a three-parameter Weibull mathematical model using continuous functions to represent discrete COVID-19 data. Subsequently, the model was applied to quantitatively analyze the characteristics for the mortality of COVID-19, including the age, sex, the length of symptom time to hospitalization time (SH), hospitalization date to death time (HD) and symptom time to death time time (SD) and others. A three-parameter mathematical model was developed by combining the reported cases in the Data Repository from the Center for Systems Science and Engineering at Johns Hopkins University and applied to estimate and analyze the characteristics for mortality of COVID-19. We found that the scale parameters of males and females were 5.85 and 5.45, respectively. Probability density functions in both males and females were negative skewness. 5% of male patients died under the age of 43.28 (44.37 for females), 50% died under 69.55 (73.25 for females), and 95% died under 86.59 (92.78 for females). The peak age of male death was 67.45 years, while that of female death was 71.10 years. The peak and median values of SH, HD and SD in male death were correspondingly 1.17, 5.18 and 10.30 days, and 4.29, 11.36 and 16.33 days, while those in female death were 1.19, 5.80 and 12.08 days, and 4.60, 12.44 and 17.67 days, respectively. The peak age of probability density in male and female deaths was 69.55 and 73.25 years, while the high point age of their mortality risk was 77.51 and 81.73 years, respectively. The mathematical model can fit and simulate the impact of various factors on IFR. From the simulation results of the model, we can intuitively find the IFR, peak age, average age and other information of each age. In terms of time factors, the mortality rate of the most susceptible population is not the highest, and the distribution of male patients is different from the distribution of females. This means that Self-protection and self-recovery in females against SARS-CoV-2 virus might be better than those of males. Males were more likely to be infected, more likely to be admitted to the ICU and more likely to die of COVID-19. Moreover, the infection fatality ration (IFR) of COVID-19 population was intrinsically linked to the infection age. Public health measures to protect vulnerable sex and age groups might be a simple and effective way to reduce IFR.

Deep clustering of small molecules at large-scale via variational autoencoder embedding and K-means.

BACKGROUND: Converting molecules into computer-interpretable features with rich molecular information is a core problem of data-driven machine learning applications in chemical and drug-related tasks. Generally speaking, there are global and local features to represent a given molecule. As most algorithms have been developed based on one type of feature, a remaining bottleneck is to combine both feature sets for advanced molecule-based machine learning analysis. Here, we explored a novel analytical framework to make embeddings of the molecular features and apply them in the clustering of a large number of small molecules. RESULTS: In this novel framework, we first introduced a principal component analysis method encoding the molecule-specific atom and bond information. We then used a variational autoencoder (AE)-based method to make embeddings of the global chemical properties and the local atom and bond features. Next, using the embeddings from the encoded local and global features, we implemented and compared several unsupervised clustering algorithms to group the molecule-specific embeddings. The number of clusters was treated as a hyper-parameter and determined by the Silhouette method. Finally, we evaluated the corresponding results using three internal indices. Applying the analysis framework to a large chemical library of more than 47,000 molecules, we successfully identified 50 molecular clusters using the K-means method with 32 embeddings based on the AE method. We visualized the clustering result via t-SNE for the overall distribution of molecules and the similarity maps for the structural analysis of randomly selected cluster-specific molecules. CONCLUSIONS: This study developed a novel analytical framework that comprises a feature engineering scheme for molecule-specific atomic and bonding features and a deep learning-based embedding strategy for different molecular features. By applying the identified embeddings, we show their usefulness for clustering a large molecule dataset. Our novel analytic algorithms can be applied to any virtual library of chemical compounds with diverse molecular structures. Hence, these tools have the potential of optimizing drug discovery, as they can decrease the number of compounds to be screened in any drug screening campaign.

Deep learning-driven prediction of drug mechanism of action from large-scale chemical-genetic interaction profiles.

MOTIVATION: Chemical-genetic interaction profiling is a genetic approach that quantifies the susceptibility of a set of mutants depleted in specific gene product(s) to a set of chemical compounds. With the recent advances in artificial intelligence, chemical-genetic interaction profiles (CGIPs) can be leveraged to predict mechanism of action of compounds. This can be achieved by using machine learning, where the data from a CGIP is fed into the machine learning platform along with the chemical descriptors to develop a chemogenetically trained model. As small molecules can be considered non-structural data, graph convolutional neural networks, which can learn from the chemical structures directly, can be used to successfully predict molecular properties. Clustering analysis, on the other hand, is a critical approach to get insights into the underlying biological relationships between the gene products in the high-dimensional chemical-genetic data. METHODS AND RESULTS: In this study, we proposed a comprehensive framework based on the large-scale chemical-genetics dataset built in Mycobacterium tuberculosis for predicting CGIPs using graph-based deep learning models. Our approach is structured into three parts. First, by matching M. tuberculosis genes with homologous genes in Escherichia coli (E. coli) according to their gene products, we grouped the genes into clusters with distinct biological functions. Second, we employed a directed message passing neural network to predict growth inhibition against M. tuberculosis gene clusters using a collection of 50,000 chemicals with the profile. We compared the performance of different baseline models and implemented multi-label tasks in binary classification frameworks. Lastly, we applied the trained model to an externally curated drug set that had experimental results against M. tuberculosis genes to examine the effectiveness of our method. Overall, we demonstrate that our approach effectively created M. tuberculosis gene clusters, and the trained classifier is able to predict activity against essential M. tuberculosis targets with high accuracy. CONCLUSION: This work provides an analytical framework for modeling large-scale chemical-genetic datasets for predicting CGIPs and generating hypothesis about mechanism of action of novel drugs. In addition, this work highlights the importance of graph-based deep neural networks in drug discovery.

Occurrence and distribution characteristics of heavy metals in the surface water of Yongding River Basin, China.

Yongding River is a vital socioeconomic zone in China in providing daily usage for humans, animals, and running of industries and agriculture. This study first provides a comparative assessment for the heavy metal pollution in the surface water from 82 estuarine locations along the basin, including the Guanting Reservoir and seven wastewater treatment plants (WWTPs). The occurrence, distribution, potential sources, and water quality relating to the detected heavy metals were addressed. Eleven typical elements (Pb, Cr, As, Cd, Sb, Ba, V, Ti, Zn, Ni, and Be) were investigated, and the results showed that all the measured concentrations were below the WHO guideline limits. Most heavy metals exhibited higher levels in the middle of Yongding River basin due to the discharge of WWTPs. Pb, Ti, Zn, and Cd in the surface water mainly originated from anthropogenic discharge, while Sb and V were mostly contributed to geogenic sources according to the principal component analysis. Three documented methods, water quality index (WQI), heavy metal pollution (HPI), and Nemerow pollution index (Pn) values, were used to evaluate the contamination monitoring of surface water. All the locations were classified as low and moderate risk except Y12, B2, and Y13 for their Pn values were higher than 1.0. The present study highlights the status of heavy metals in Yongding River basin which is helpful in providing fundamental data for assessment of water quality and the effective protection for Yongding River basin in the future.

A practical method to screen and identify functioning biomarkers in nasopharyngeal carcinoma.

Nasopharyngeal carcinoma (NPC) is a rare malignancy, with the unique geographical and ethnically characteristics of distribution. Gene chip and bioinformatics have been employed to reveal regulatory mechanisms in current functional genomics. However, a practical solution addressing the unresolved aspects of microarray data processing and analysis have been long pursuit. This study developed a new method to improve the accuracy of identifying key biomarkers, namely Unit Gamma Measurement (UGM), accounting for multiple hypotheses test statistics distribution, which could reduce the dependency problem. Three mRNA expression profile of NPC were selected to feed UGM. Differentially expressed genes (DEGs) were identified with UGM and hub genes were derived from them to explore their association with NPC using functional enrichment and pathway analysis. 47 potential DEGs were identified by UGM from the 3 selected datasets, and affluent in cysteine-type endopeptidase inhibitor activity, cilium movement, extracellular exosome etc. also participate in ECM-receptor interaction, chemical carcinogenesis, TNF signaling pathway, small cell lung cancer and mismatch repair pathway. Down-regulation of CAPS and WFDC2 can prolongation of the overall survival periods in the patients. ARMC4, SERPINB3, MUC4 etc. have a close relationship with NPC. The UGM is a practical method to identify NPC-associated genes and biomarkers.

Distribution, toxicity load, and risk assessment of dissolved metal in surface and overlying water at the Xiangjiang River in southern China.

Metal pollution in drinking water source has been under scrutiny as it seriously affects human health. This work examined 12 dissolved metals in the surface and overlying water of the Xiangjiang River, an important drinking water source in southern China, and characterized their distribution, identified their possible sources, assessed their toxicity load, and determined their potential ecological and health risk. No significant difference was found in the metal concentration between surface and overlying water. The average metal concentration fell in the order of Mg > Mn > Ba > Fe > Zn > As > Sb > Ni > Cd > V > Cr > Co, and all was lower than the safety threshold in the drinking water guideline of China. Anthropogenic activities were found to be the main source of metals from correlation analysis, principal component analysis (PCA), and cluster analysis (CA). According to the total heavy metal toxicity load (HMTL), 98.20%, 71.54%, 68.88%, and 7.97% of As, Cd, Sb, and Mn should be removed from the surface water to ensure safety. Most water samples from the surveyed area were found to have high ecological risk as was measured by the ecological risk index (RI). Health risk assessment showed that children are more susceptible than adults to the non-carcinogenic risk of dissolved metals, and the potential carcinogenic risk (CR) of As and Cd should be addressed. The results provide guidance for controlling the metal pollution of the Xiangjiang River and improving its quality as a drinking water source.

UGM: a more stable procedure for large-scale multiple testing problems, new solutions to identify oncogene.

Variations of gene expression levels play an important role in tumors. There are numerous methods to identify differentially expressed genes in high-throughput sequencing. Several algorithms endeavor to identify distinctive genetic patterns susceptable to particular diseases. Although these processes have been proved successful, the probability that the number of non-differentially expressed genes measured by false discovery rate (FDR) has a large standard deviation, and the misidentification rate (type I error) grows rapidly when the number of genes to be detected become larger. In this study we developed a new method, Unit Gamma Measurement (UGM), accounting for multiple hypotheses test statistics distribution, which could reduce the dependency problem. Simulated expression profile data and breast cancer RNA-Seq data were utilized to testify the accuracy of UGM. The results show that the number of non-differentially expressed genes identified by the UGM is very close to the real-evidence data, and the UGM also has a smaller standard error, range, quartile range and RMS error. In addition, the UGM can be used to screen many breast cancer-associated genes, such as BRCA1, BRCA2, PTEN, BRIP1, etc., provides better accuracy, robustness and efficiency, the method of identification differentially expressed genes in high-throughput sequencing.

High contamination, bioaccumulation and risk assessment of perfluoroalkyl substances in multiple environmental media at the Baiyangdian Lake.

The contamination of perfluoroalkyl substances (PFASs) in the Baiyangdian Lake has exacerbated readily since 2008. This study analyzed the perfluoroalkyl carboxylic acids (PFCAs) and perfluoroalkane sulfonic acids (PFSAs) in the surface water, sediment, and fish of the Baiyangdian Lake. In the surface water, the total concentration of PFASs ranged in 1193-3462 ng L-1 (mean 1734 ng L-1) in the rainy season and 469-1724 ng L-1 (mean 876 ng L-1) in the dry season. The total concentration of PFASs in the sediment ranged in 1.97-13.3 ng g-1 (mean 6.53 ng g-1). It was found that PFCAs and PFSAs with longer chains were more easily adsorbed in the sediment. Among the collected fish samples, the enrichment of PFASs in the tissues fell in the order of liver > cheek > muscle. For the muscle, stomach, and liver tissues of the fish samples, significant correlations existed between the δ15N values and the concentration of perfluorooctane sulfonic acid (PFOS). The contents of PFOS and perfluorooctanoic acid (PFOA) in the fish were not at a level high enough to significantly risk human health.

Risk assessment of organochlorine pesticides in drinking water source of the Yangtze River.

Organochlorine pesticides have been banned for many years, but the residual trace amount of organochlorine in water may still pose ecotoxicological risk. Meanwhile, the potential risk of organochlorine pesticides released from sediments, especially into drinking water sources, is receiving increasing attention. The present study assessed the pollution and potential risk of drinking water sources along the midstream and downstream Yangtze River. Residues of organochlorine pesticides (OCPs) in water, suspended particle matter (SPM), and sediment were evaluated with isotope dilution HRGC/HRMS. The results indicated that OCPs in water, SPM, and sediment ranged in 0.52-92.97 ng/L, 0.10-4.10 ng/L, and 0.038-11.36 ng/g, respectively. The predominant OCPs in water, SPM, and sediment were ß-HCH, p,p'-DDE and PeCB. At site Y1, 8, 13, 18, ß-HCH has a higher proportion in sediment samples, while, α-HCH has a higher proportion in SPM samples. The industrial use of HCHs in the history was the main HCHs source for most water and sediment samples, which indicated an absence of fresh inputs of industrial HCHs. Meanwhile, the abundance of p,p'-DDE in water, sediment and SPM samples could be attributed to long-term aerobic degradation of DDTs. The values of ffsw of HCHs, DDTs and PeCB indicate the transfer from water to sediment. Risk assessment showed that HCHs and DDTs posed low ecotoxicological risk to the Yangtze River.