Porcine deltacoronavirus nucleocapsid protein antagonizes JAK-STAT signaling pathway by targeting STAT1 through KPNA2 degradation.

Porcine deltacoronavirus (PDCoV) is an enteric pathogenic coronavirus that causes acute and severe watery diarrhea in piglets and has the ability of cross-species transmission, posing a great threat to swine production and public health. The interferon (IFN)-mediated signal transduction represents an important component of virus-host interactions and plays an essential role in regulating viral infection. Previous studies have suggested that multifunctional viral proteins encoded by coronaviruses antagonize the production of IFN via various means. However, the function of these viral proteins in regulating IFN-mediated signaling pathways is largely unknown. In this study, we demonstrated that PDCoV and its encoded nucleocapsid (N) protein antagonize type I IFN-mediated JAK-STAT signaling pathway. We identified that PDCoV infection stimulated but delayed the production of IFN-stimulated genes (ISGs). In addition, PDCoV inhibited JAK-STAT signal transduction by targeting the nuclear translocation of STAT1 and ISGF3 formation. Further evidence showed that PDCoV N is the essential protein involved in the inhibition of type I IFN signaling by targeting STAT1 nuclear translocation via its C-terminal domain. Mechanistically, PDCoV N targets STAT1 by interacting with it and subsequently inhibiting its nuclear translocation. Furthermore, PDCoV N inhibits STAT1 nuclear translocation by specifically targeting KPNA2 degradation through the lysosomal pathway, thereby inhibiting the activation of downstream sensors in the JAK-STAT signaling pathway. Taken together, our results reveal a novel mechanism by which PDCoV N interferes with the host antiviral response.IMPORTANCEPorcine deltacoronavirus (PDCoV) is a novel enteropathogenic coronavirus that receives increased attention and seriously threatens the pig industry and public health. Understanding the underlying mechanism of PDCoV evading the host defense during infection is essential for developing targeted drugs and effective vaccines against PDCoV. This study demonstrated that PDCoV and its encoded nucleocapsid (N) protein antagonize type I interferon signaling by targeting STAT1, which is a crucial signal sensor in the JAK-STAT signaling pathway. Further experiments suggested that PDCoV N-mediated inhibition of the STAT1 nuclear translocation involves the degradation of KPNA2, and the lysosome plays a role in KPNA2 degradation. This study provides new insights into the regulation of PDCoV N in the JAK-STAT signaling pathway and reveals a novel mechanism by which PDCoV evades the host antiviral response. The novel findings may guide us to discover new therapeutic targets and develop live attenuated vaccines for PDCoV infection.

Analysis of Factors Affecting Concentrations and Concentration-To-Dose Ratios of Trazodone.

BACKGROUND: Trazodone is prescribed for several clinical conditions. Multiple factors may affect trazodone to reach its therapeutic reference range. The concentration-to-dose (C/D) ratio can be used to facilitate the therapeutic drug monitoring of trazodone. The study aimed to investigate factors on the concentrations and C/D ratio of trazodone. METHODS: This study analyzed the therapeutic drug monitoring electronic case information of inpatients in the First Hospital of Hebei Medical University from October 2021 to July 2023. Factors that could affect the concentrations and C/D ratio of trazodone were analyzed, including body mass index, sex, age, smoking, drinking, drug manufacturers, and concomitant drugs. RESULTS: A total of 255 patients were analyzed. The mean age was 52.44 years, and 142 (55.69%) were women. The mean dose of trazodone was 115.29 mg. The mean concentration of trazodone was 748.28 ng/mL, which was in the therapeutic reference range (700-1000 ng/mL). 50.20% of patients reached the reference range, and some patients (36.86%) had concentrations below the reference range. The mean C/D ratio of trazodone was 6.76 (ng/mL)/(mg/d). A significant positive correlation was found between daily dose and trazodone concentrations (r 2 = 0.2885, P < 0.001). Trazodone concentrations were significantly affected by dosage, sex, smoking, drinking, and concomitant drugs of duloxetine or fluoxetine. After dosage emendation, besides the above factors, it was influenced by age ( P < 0.05, P < 0.01, or P < 0.001). CONCLUSIONS: This study identified factors affecting trazodone concentrations and C/D ratio. The results can help clinicians closely monitor patients on trazodone therapy and maintain concentrations within the reference range.

Obscuring effect of income inequality and moderating role of financial literacy in the relationship between digital finance and China's household carbon emissions.

Households have emerged as one of the primary sources for carbon emissions in China, thus posing challenges to the "dual carbon" objectives. Digital finance, an emergent form of industry that fused advanced technology with financial services, had a pronounced impact on household carbon emissions stemming from daily consumption. However, the mechanisms driving this impact have not been adequately examined. Based on micro-level household survey data across 25 Chinese provinces from 2012, 2014, 2016, and 2018, the study identified the chief channels via which digital finance affected household carbon emissions, deriving several key findings. First, digital finance augmented household carbon emissions, presenting a significant negative impact on the climate. Second, due to the existence of "digital divide" between rural and urban areas, the impact of digital finance was more subdued in rural areas. Additionally, the effects of digital finance were more pronounced in the affluent eastern provinces. Third, income mobility obscured the positive relationship between digital finance and household carbon emissions. This is primarily attributed to the urban-rural divide in China; taking into account that urban-to-rural transfers make income distribution more equitable, there is a counterintuitive drop in per capita consumption, thereby suppressing consumption-related carbon emissions. This presented the conundrum of "income distribution equality-consumption negativity". Finally, financial literacy was identified as a crucial positive moderating role, enabling households with high financial literacy to harness the dividends of digital finance, thereby engaging in more diversified consumption activities and intensifying the negative impact of digital finance on carbon emissions. The findings reinforced the pivotal role of digital finance in bolstering efforts to combat climate change and ensuring environmentally-responsible economic advancements.

TFRegNCI: Interpretable Noncovalent Interaction Correction Multimodal Based on Transformer Encoder Fusion.

The interpretability is an important issue for end-to-end learning models. Motivated by computer vision algorithms, an interpretable noncovalent interaction (NCI) correction multimodal (TFRegNCI) is proposed for NCI prediction. TFRegNCI is based on RegNet feature extraction and a transformer encoder fusion strategy. RegNet is a network design paradigm that mainly focuses on local features. Meanwhile, the Vision Transformer is also leveraged for feature extraction, because it can capture global features better than RegNet while lowering the computational cost. Using a transformer encoder as the fusion strategy rather than multilayer perceptron can enhance model performance, due to its emphasis on important features with less parameters. Therefore, the proposed TFRegNCI achieved high accurate prediction (mean absolute error of ∼0.1 kcal/mol) comparing with the coupled cluster single double (triple) (CCSD(T)) benchmark. To further improve the model efficiency, TFRegNCI applies two-dimensional (2D) inputs transformed from three-dimensional (3D) electron density cubes, which saves time (30%), while the model accuracy remains. To improve model interpretability, a visualization module, Gradient-weighted Regression Activation Mapping (Grad-RAM) has been embedded. Grad-RAM is promoted from the classification algorithm, Gradient-weighted Class Activation Mapping, to perform feature visualization for the regression task. With Grad-RAM, the visual location map for features in deep learning models can be displayed. The feature map visualizations suggest that the 2D model has the similar performance as the 3D model, because of equally effective feature extractions from electron density. Moreover, the valid feature region on the location map by the 3D model is consistent with the NCIPLOT NCI isosurface. It is confirmed that the model does extract significant features related to the NCI interaction. The interpretable analyses are carried out through molecular orbital contribution on effective features. Thereby, the proposed model is likely to be a promising tool to reveal some essential information on NCIs, with regard to the level of electronic theory.

Efficient prediction for high precision CO-N2 potential energy surface by stacking ensemble DNN.

High-dimensional potential energy surface (PES) for van der Waals systems with spectroscopic accuracy, is of great importance for quantum dynamics and an extremely challenge job. CO-N2 is a typical van der Waals system and its high-precision PES may help elucidate weak interaction mechanisms. Taking CO-N2 potential energies calculated by CCSD(T)-F12b/aug-cc-pVQZ as the benchmark, we establish an accurate, robust, and efficient machine learning model by using only four molecular structure descriptors based on 7966 benchmark potential energies. The highest accuracy is obtained by a stacking ensemble DNN (SeDNN). Its evaluation parameters MAE, RMSE, and R2 reach 0.096, 0.163, 0.9999 cm-1 , respectively, and the spectroscopic accuracy for vibration spectrum is achieved with predicted PES, which shows SeDNN superior goodness-of-fit and prediction performance. An elaborated PES with the reported global minimum has been predicted with the model, which perfectly reproduces CCSD(T) potential energies and the analytical MLR PES [PCCP, 2018, 20, 2036]. The critical points (global minimum, TSI, TSII, and their barriers), potential curve, and entire PES profile are remarkably consistent with CCSD(T) calculations. To further improve the usability of constructing PESs in practice, the size of the training set (energy points) for the model is reduced to 50%, 30%, and 20% of the database, respectively. The results show that even training with the smallest training set (1593 points), the PES only differs 2.555 cm-1 with the analytic MLR PES. Therefore, the proposed SeDNN is promisingly an alternative efficient tool to construct subtle PES for van der Waals systems.

DeepNCI: DFT Noncovalent Interaction Correction with Transferable Multimodal Three-Dimensional Convolutional Neural Networks.

A multimodal deep learning model, DeepNCI, is proposed for improving noncovalent interactions (NCIs) calculated via density functional theory (DFT). DeepNCI is composed of a three-dimensional convolutional neural network (3D CNN) for abstracting critical and comprehensive features from 3D electron density, and a neural network for modeling one-dimensional quantum chemical properties. By merging features from two networks, DeepNCI is able to reduce the root-mean-square error of DFT-calculated NCI from 1.19 kcal/mol to ∼0.2 kcal/mol for a NCI molecular database (>1000 molecules). The representativeness of the joint features can be visualized by t-distributed stochastic neighbor embedding (t-SNE), where they can distinguish categorized NCI systems quite well. Therefore, the fused model performs better than its component networks. In addition, the 3D CNN takes electron density as inputs that are in the same range, despite the size of molecular systems, so it can promote model applicability and transferability. To clarify the applicability of DeepNCI, an application domain (AD) has been defined with merged features using the K-nearest-neighbor method. The calculations for external test sets are shown that AD can properly monitor the reliability for a prediction. The model transferability is tested with a small database of homolysis bond dissociation energy including only dozens of samples. With NCI database pretrained parameters, the same or better performance than the reported results is achieved by transfer learning. This suggests that the DeepNCI model is transferable and it may transfer to other relative tasks, which possibly can resolve some small sampling problems. The source code of DeepNCI can be freely accessed at https://github.com/wenzelee/DeepNCI.

Responses of Hydrocharis dubia (Bl.) Backer and Trapa bispinosa roxb. to tetracycline exposure.

The presence of tetracycline is ubiquitous and has adverse effects on aquatic systems. A hydroponic experiment was conducted to investigate the ecological sensitivity of Hydrocharis dubia (Bl.) Backer and Trapa bispinosa Roxb. Exposed to different concentrations of tetracycline (0, 0.1, 1, 10, 30 and 50 mg/L) for one day (1D) and 14 days (14D). The results showed that after 1D of tetracycline exposure, the physiological indices of H. dubia had no remarkable change except for proline which was significantly stimulated under 0.1 mg/L tetracycline. For T. bispinosa, guaiacol peroxidase (POD), polyphenol oxidase (PPO) and ascorbate peroxidase (APX) activity and protein and proline content were notably promoted under different concentrations of tetracycline, but PPO activity was significantly decreased in 50 mg/L. After 14D, tetracycline caused no harm to the growth and protein content of H. dubia, but negatively influenced lipid peroxidation product and chlorophyll content in H. dubia under high tetracycline concentrations. Superoxide dismutase (SOD) and POD activity of H. dubia significantly increased at high tetracycline concentrations, while catalase (CAT) and PPO activity significantly decreased. APX activity in H. dubia increased with tetracycline concentrations at low tetracycline concentrations. For T. bispinosa, high concentrations of tetracycline application significantly inhibited its growth and the content of protein and chlorophyll. SOD, POD, CAT, and PPO activity of T. bispinosa were induced under different concentrations of tetracycline and no lipid peroxidation was observed. APX activity in T. bispinosa was significantly inhibited at high tetracycline concentrations. The results suggest that tetracycline can cause oxidative damage in H. dubia but harm the metabolism process of T. bispinosa without inducing oxidative damage. Overall, the sensitivity of T. bispinosa exposed to tetracycline exposure is higher than that of H. dubia.

Using Supervised Machine Learning to Classify Real Alerts and Artifact in Online Multisignal Vital Sign Monitoring Data.

OBJECTIVE: The use of machine-learning algorithms to classify alerts as real or artifacts in online noninvasive vital sign data streams to reduce alarm fatigue and missed true instability. DESIGN: Observational cohort study. SETTING: Twenty-four-bed trauma step-down unit. PATIENTS: Two thousand one hundred fifty-three patients. INTERVENTION: Noninvasive vital sign monitoring data (heart rate, respiratory rate, peripheral oximetry) recorded on all admissions at 1/20 Hz, and noninvasive blood pressure less frequently, and partitioned data into training/validation (294 admissions; 22,980 monitoring hours) and test sets (2,057 admissions; 156,177 monitoring hours). Alerts were vital sign deviations beyond stability thresholds. A four-member expert committee annotated a subset of alerts (576 in training/validation set, 397 in test set) as real or artifact selected by active learning, upon which we trained machine-learning algorithms. The best model was evaluated on test set alerts to enact online alert classification over time. MEASUREMENTS AND MAIN RESULTS: The Random Forest model discriminated between real and artifact as the alerts evolved online in the test set with area under the curve performance of 0.79 (95% CI, 0.67-0.93) for peripheral oximetry at the instant the vital sign first crossed threshold and increased to 0.87 (95% CI, 0.71-0.95) at 3 minutes into the alerting period. Blood pressure area under the curve started at 0.77 (95% CI, 0.64-0.95) and increased to 0.87 (95% CI, 0.71-0.98), whereas respiratory rate area under the curve started at 0.85 (95% CI, 0.77-0.95) and increased to 0.97 (95% CI, 0.94-1.00). Heart rate alerts were too few for model development. CONCLUSIONS: Machine-learning models can discern clinically relevant peripheral oximetry, blood pressure, and respiratory rate alerts from artifacts in an online monitoring dataset (area under the curve > 0.87).

Machine learning-based prediction of sertraline concentration in patients with depression through therapeutic drug monitoring.

Background: Sertraline is a commonly employed antidepressant in clinical practice. In order to control the plasma concentration of sertraline within the therapeutic window to achieve the best effect and avoid adverse reactions, a personalized model to predict sertraline concentration is necessary. Aims: This study aimed to establish a personalized medication model for patients with depression receiving sertraline based on machine learning to provide a reference for clinicians to formulate drug regimens. Methods: A total of 415 patients with 496 samples of sertraline concentration from December 2019 to July 2022 at the First Hospital of Hebei Medical University were collected as the dataset. Nine different algorithms, namely, XGBoost, LightGBM, CatBoost, random forest, GBDT, SVM, lasso regression, ANN, and TabNet, were used for modeling to compare the model abilities to predict sertraline concentration. Results: XGBoost was chosen to establish the personalized medication model with the best performance (R 2 = 0.63). Five important variables, namely, sertraline dose, alanine transaminase, aspartate transaminase, uric acid, and sex, were shown to be correlated with sertraline concentration. The model prediction accuracy of sertraline concentration in the therapeutic window was 62.5%. Conclusion: In conclusion, the personalized medication model of sertraline for patients with depression based on XGBoost had good predictive ability, which provides guidance for clinicians in proposing an optimal medication regimen.

Artificial intelligence-based model for dose prediction of sertraline in adolescents: a real-world study.

BACKGROUND: Variability exists in sertraline pharmacokinetic parameters in individuals, especially obvious in adolescents. We aimed to establish an individualized dosing model of sertraline for adolescents with depression based on artificial intelligence (AI) techniques. METHODS: Data were collected from 258 adolescent patients treated at the First Hospital of Hebei Medical University between December 2019 to July 2022. Nine different algorithms were used for modeling to compare the prediction abilities on sertraline daily dose, including XGBoost, LGBM, CatBoost, GBDT, SVM, ANN, TabNet, KNN, and DT. Performance of four dose subgroups (50 mg, 100 mg, 150 mg, and 200 mg) were analyzed. RESULTS: CatBoost was chosen to establish the individualized medication model with the best performance. Six important variables were found to be correlated with sertraline dose, including plasma concentration, PLT, MPV, GL, A/G, and LDH. The ROC curve and confusion matrix exhibited the good prediction performance of CatBoost model in four dose subgroups (the AUC of 50 mg, 100 mg, 150 mg, and 200 mg were 0.93, 0.81, 0.93, and 0.93, respectively). CONCLUSION: The AI-based dose prediction model of sertraline in adolescents with depression had a good prediction ability, which provides guidance for clinicians to propose the optimal medication regimen.

Separate and combined effects of cadmium (Cd) and nonylphenol (NP) on growth and antioxidative enzymes in Hydrocharis dubia (Bl.) Backer.

Cadmium (Cd) is considered a priority pollutant, and nonylphenol (NP) is a common organic pollutant in water environments. However, the ecological risks of combined Cd and NP pollution have not been fully elucidated. In this study, the effects of Cd, NP, and Cd-NP on the growth and physiology of Hydrocharis dubia (Bl.) Backer were studied. The results indicated that Cd-NP joint toxicity is concentration-dependent. The joint toxicity of Cd and NP on H. dubia was antagonistic when the concentrations of Cd + NP were 0.01 + 0.1/1 mg/L. At 0.5 + 0.1/1 mg/L, Cd and NP had a strong synergistic effect on H. dubia. In addition, plant growth was significantly inhibited, and the chlorophyll contents were significantly reduced under Cd, NP, or Cd-NP exposure. The plant's antioxidant enzyme system was destroyed. The activities of superoxide dismutase (SOD) and catalase (CAT) were significantly decreased under NP-only exposure. The activity of SOD was significantly decreased under Cd-only and under joint exposure. Compound pollution exceeded the oxidative defense capacity of the plants, so the H2O2 content increased significantly. Our results indicated that the ecotoxicity of NP combined with Cd may be exacerbated in aquatic environments and cause obvious damage to H. dubia.

Hepatoprotective effect and structural analysis of Hedysarum polysaccharides in vivo and in vitro.

The crude Hedysarum polysaccharides (HPS: HPS-50 and HPS-80) obtained from Radix Hedysari exhibited great pharmacological activities in our previous research. This study investigated the effects of HPS on lipopolysaccharide (LPS)/D-galactosamine (D-GalN)-induced acute liver injury (ALI) in mice and LPS-induced injury in LO2 cells, as well as the relationship between structural characteristics and hepatoprotective activities. The in vivo results showed that compared with HPS-80, HPS-50 showed stronger hepatoprotection, which improved histopathological changes to normal levels. HPS-50 significantly decreased the levels of ALT, AST, MPO, and MDA, increased the activities of SOD, CAT, and GSH, and suppressed the LPS/D-GalN-triggered production of TNF-α, IL-1ß, and IL-6 (p < .05). The results in vitro showed that HPS-50-P (HPS-50-1, HPS-50-2, and HPS-50-3) purified from HPS-50 played significant protective roles against LPS-induced injury in LO2 cells by reducing cell apoptosis and relieving cell cycle arrest. HPS-50-2 restored the percentage of normal cells from 54.8% to 94.7%, and reduced the S phase cells from 59.40% to 47.05% (p < .01). By analyzing the structure of HPS-50-P, including monosaccharide composition, molecular weight, chain conformation, and surface morphology, we speculated that the best protective effect of HPS-50-2 might be attributed to its beta configuration, highest molecular weight, and high glucose and galactose contents. These findings indicate that HPS-50 might be a promising source of functional foods for the protection and prevention of ALI. PRACTICAL APPLICATIONS: In this study, the protective effect of HPS on ALI was evaluated from multiple perspectives, and HPS-50-2 was screened as a potential active ingredient. This study has two practical applications. First, it provides a new way to improve ALI, and a new option for patients to prevent and treat ALI. Second, this work also complements the pharmacological activity of Radix Hedysari and provides a basis for the development of Radix Hedysari as a functional food.

Green synthesis, characterization of Radix Hedysari-mediated silver nanoparticles and their use for sensitive colorimetric detection of Pb2+ in the Yellow River medium.

In this study, a safe, rapid, and environment-friendly green synthesis of silver nanoparticles using the alcohol extract of Radix Hedysari (RH-AgNPs) was developed, the alcohol extract of Radix Hedysari (RH) acted as the reducing agent, stabilizer, and modifier. The main components of RH were determined using high-performance liquid chromatography (HPLC). The particle size and morphology of RH-AgNPs were optimized and characterized by a series of techniques. The size distribution, zeta potential, element distribution, and crystalline nature of RH-AgNPs were all determined. It was indicated that RH-AgNPs showed great sensitivity for lead ion (Pb2+) detection with a limit of detection (LOD) of 1.5 µM with a wide range of 10-500 µM. The selectivity was also explored for common metal ions. RH-AgNPs were then applied to the detection of Pb2+ in spiked Yellow River samples, and the possible mechanism is based on the crosslinking reaction between the hydroxide radical, carboxylate radical and Pb2+.

Structural characterization of a polysaccharide from Radix Hedysari and its protective effects against H2O2-induced injury in human gastric epithelium cells.

The gastroprotective effects of polysaccharides had become a hot topic in the field of functional polysaccharides research. Three polysaccharides, namely HPS-80-1, HPS-80-2, and HPS-80-3 were purified by DEAE-52 column chromatography. The thermodynamic characteristics, scanning electron microscopy, and Congo red experimental results of the above polysaccharides were greatly distinctive. Then a mature GES-1 oxidative stress cell model induced by H2O2 was established to screen out subsequent research subjects. It turned out that HPS-80-1 had a desirable protective effect, which was confirmed by analyses of cell cycle & apoptosis, and oxidative stress-related factors in the cell culture media, and so on. Furthermore, Structural features demonstrated that the backbone of HPS-80-1 appeared to mainly consist of →4)-α-D-Glcp-(1→, →4,6)-ß-L-Glcp-(1→, and →6)-α-D-Galp-(1→, with branches at O-1, O-4, and O-6 position consisting of →2,4)-ß-D-Rhap-(1→, →1)-α-D-Galp-(4→, and →3,4)-α-D-Manp-(1→. It was speculated that the excellent gastric mucosal protective activity of HPS-80-1 may be due to the high amount of glucose in the backbone. In addition, it was also related to the anti-inflammatory activity and antioxidant bases such as (1 â†’ 4)-Glcp and (1 â†’ 6)-Galp in the structure of HPS-80-1. These findings provide a scientific basis for further utilization of polysaccharides from Radix Hedysari.

The amelioration of ulcerative colitis induced by Dinitrobenzenesulfonic acid with Radix Hedysari.

Ulcerative colitis (UC) is a chronic inflammatory disease with an unknown precise etiology. This study proves that Radix Hedysari (RH) ameliorates UC. Four RH extracts were used to ameliorate UC induced by 2,4-Dinitrobenzenesulfonic acid by 7 days intervention in agreement to preliminary studies. Compared to treatment with RH extracts, the RH ethanol extract (EE) was found to be more effective in ameliorating UC. With EE, the DAI were significantly decreased. Macroscopic and histopathological assessments suggest that the colon mucosa was repaired, the organizational structure of the colon had been rebuilt. The levels of MPO, TNF-α, IL-1ß, and MDA were significantly decreased (p < .01), the levels of T-SOD and CAT were significantly increased (p < .01). Moreover, the compounds in EE were analyzed by HPLC. The results show that EE can ameliorate UC, and its anti-inflammatory capability probably plays an important role. RH can act as a functional food and ameliorate UC. PRACTICAL APPLICATIONS: In this work, the ameliorative effect of RH on UC was evaluated from multiple angles. There are two practical applications of this work. On the one hand, a new approach to ameliorating UC is provided by this work. In addition, UC patients have a new option for improving their symptoms. On the other hand, this work also provides information on how best to process RH for therapeutic use. In addition, we can utilize some compounds of RH that were once considered useless and reduce the waste of natural resources.

Learning temporal rules to forecast instability in continuously monitored patients.

Inductive machine learning, and in particular extraction of association rules from data, has been successfully used in multiple application domains, such as market basket analysis, disease prognosis, fraud detection, and protein sequencing. The appeal of rule extraction techniques stems from their ability to handle intricate problems yet produce models based on rules that can be comprehended by humans, and are therefore more transparent. Human comprehension is a factor that may improve adoption and use of data-driven decision support systems clinically via face validity. In this work, we explore whether we can reliably and informatively forecast cardiorespiratory instability (CRI) in step-down unit (SDU) patients utilizing data from continuous monitoring of physiologic vital sign (VS) measurements. We use a temporal association rule extraction technique in conjunction with a rule fusion protocol to learn how to forecast CRI in continuously monitored patients. We detail our approach and present and discuss encouraging empirical results obtained using continuous multivariate VS data from the bedside monitors of 297 SDU patients spanning 29 346 hours (3.35 patient-years) of observation. We present example rules that have been learned from data to illustrate potential benefits of comprehensibility of the extracted models, and we analyze the empirical utility of each VS as a potential leading indicator of an impending CRI event.