Retrosynthesis Zero: Self-Improving Global Synthesis Planning Using Reinforcement Learning.

The field of computer-aided synthesis planning (CASP) has witnessed significant growth in recent years. Still, many CASP programs rely on large data sets to train neural networks, resulting in limitations due to the data quality and prior knowledge from chemists. In response, we propose Retrosynthesis Zero (ReSynZ), a reaction template-based method that combines Monte Carlo Tree Search with reinforcement learning inspired by AlphaGo Zero. Unlike other single-step reaction template-based CASP methods, ReSynZ takes complete synthesis paths for complex molecules, determined by reaction rules, as input for training the neural network. ReSynZ enables neural networks trained with relatively small reaction data sets (tens of thousands of data) to generate multiple synthesis pathways for a target molecule and suggest possible reaction conditions. On multiple data sets of molecular retrosynthesis, ReSynZ demonstrates excellent predictive performance compared to existing algorithms. The advantages, such as self-improving model features, flexible reward settings, the potential to surpass human limitations in chemical synthesis route planning, and others, make ReSynZ a valuable tool in chemical synthesis design.

Perceived Worries and Spirituality: A Mixed Methods Study of the Primary Caregiver Well-Being of Orphan and Vulnerable Children in Ethiopia.

This study investigates the well-being of primary caregivers responsible for orphaned and vulnerable children. Well-being is defined as overall wellness, happiness, and satisfaction. Through mixed methods case studies and purposive sampling, we analyzed data from the Ziway Food for the Hungry Ethiopia program in 2017. Our explanatory analytic approach highlighted issues including resource constraints, chronic illnesses, and community challenges faced by the respondents. Nonetheless, spiritual well-being emerged as a crucial factor for their coping mechanisms. The findings underscore that critical well-being deficiencies require immediate attention. Strategies should prioritize financial and emotional support, emphasizing community capital to enhance the well-being of primary caregivers.

Experimental and Numerical Investigation of Prepreg-RTM Co-Curing Molding Composite Bolted T-Joint under Bending Load.

A set of polymer composite bolted T-joints with a novel configuration consisting of an internal skeleton and external skin was fabricated using a prepreg-RTM co-curing molding process. Experiments were conducted to study their mechanical properties under a bending load. A finite element model with a polymer resin area between the skin and skeleton was established and verified by the experimental results. Then, the damage propagation process and failure mechanism of the joint and the influence of three factors related to the layer characteristics of the skin and skeleton were investigated by the validated models. The results show that the bending stiffness and the yield limit load of the novel composite T-joint are 0.81 times and 1.65 times that of the 2A12 aluminum T-joint, respectively, while at only 55.4% of its weight. The damage of the joint is initiated within the resin area and leads to the degradation of the joint's bending performance. The preferred stacking sequence of the skeleton is [0/+45/90/-45]ns when primarily subjected to bending loads. The decrease in the bending performance is within 5% of the inclining angle of the skeleton, less than 12 degrees. The more 90° layers in the skin, the better the bending performance of the joints, while the more 0° layers, the poorer the bending performance.

Compact single-shot soft X-ray photon spectrometer for free-electron laser diagnostics.

The photon spectrum from free-electron laser (FEL) light sources offers valuable information in time-resolved experiments and machine optimization in the spectral and temporal domains. We have developed a compact single-shot photon spectrometer to diagnose soft X-ray spectra. The spectrometer consists of an array of off-axis Fresnel zone plates (FZP) that act as transmission-imaging gratings, a Ce:YAG scintillator, and a microscope objective to image the scintillation target onto a two-dimensional imaging detector. This spectrometer operates in segmented energy ranges which covers tens of electronvolts for each absorption edge associated with several atomic constituents: carbon, nitrogen, oxygen, and neon. The spectrometer's performance is demonstrated at a repetition rate of 120 Hz, but our detection scheme can be easily extended to 200 kHz spectral collection by employing a fast complementary metal oxide semiconductor (CMOS) line-scan camera to detect the light from the scintillator. This compact photon spectrometer provides an opportunity for monitoring the spectrum downstream of an endstation in a limited space environment with sub-electronvolt energy resolution.

Prediction on X-ray output of free electron laser based on artificial neural networks.

Knowledge of x-ray free electron lasers' (XFELs) pulse characteristics delivered to a sample is crucial for ensuring high-quality x-rays for scientific experiments. XFELs' self-amplified spontaneous emission process causes spatial and spectral variations in x-ray pulses entering a sample, which leads to measurement uncertainties for experiments relying on multiple XFEL pulses. Accurate in-situ measurements of x-ray wavefront and energy spectrum incident upon a sample poses challenges. Here we address this by developing a virtual diagnostics framework using an artificial neural network (ANN) to predict x-ray photon beam properties from electron beam properties. We recorded XFEL electron parameters while adjusting the accelerator's configurations and measured the resulting x-ray wavefront and energy spectrum shot-to-shot. Training the ANN with this data enables effective prediction of single-shot or average x-ray beam output based on XFEL undulator and electron parameters. This demonstrates the potential of utilizing ANNs for virtual diagnostics linking XFEL electron and photon beam properties.

Clinical characteristics and long-term neurodevelopmental outcomes of leukomalacia in preterm infants and term infants: a cohort study.

BACKGROUND: Leukomalacia is a serious form of neonatal brain injury that often leads to neurodevelopmental impairment, and studies on neonatal leukomalacia and its long-term outcomes are lacking. The aim of this study was to analyze the clinical manifestations, imaging features, and long-term neurodevelopmental outcomes in preterm infants and term infants with leukomalacia. METHODS: Newborns diagnosed with leukomalacia by head magnetic resonance imaging (MRI) and who were admitted to intensive care units from January 2015 to June 2020 were enrolled. All infants were followed up to June 2022 (2-7 years old), and their neurodevelopmental outcomes were evaluated. The clinical data and long- term outcomes of preterm infants and term infants was analyzed by Chi-square tests. RESULTS: A total of 218 surviving infants with leukomalacia including 114 preterm infants and 104 term infants completed the follow-up. The major typesof leukomalacia on MRI were periventricular leukomalacia in the preterm group and subcortical cystic leukomalacia in the term group, respectively (χ2 = 55.166; p < 0.001). When followed up to 2-7 years old, the incidence of neurodevelopmental impairment in the preterm group and term group was not significantly different (χ2 = 0.917; p = 0.338). However, the incidence of cerebral palsy (CP) in the preterm group was significantly higher (χ2 = 4.896; p = 0.027), while the incidence of intellectual disability (ID) (χ2 = 9.445; p = 0.002), epilepsy (EP) (χ2 = 23.049; p < 0.001), and CP combined with ID andEP (χ2 = 4.122; p = 0.042) was significantly lower than that in the term group. CONCLUSIONS: Periventricular leukomalacia mainly occurred in preterm infants while subcortical cystic leukomalacia was commonly seen in term infants. Although the long-term neurodevelopmental outcomes of leukomalacia were both poor, preterm infants were more prone to CP, while term infants were more prone to ID, EP, and the combination of CP with ID and EP.

Effects of oxidative stress on hepatic encephalopathy pathogenesis in mice.

Oxidative stress plays a crucial role in the pathogenesis of hepatic encephalopathy (HE), but the mechanism remains unclear. GABAergic neurons in substantia nigra pars reticulata (SNr) contribute to the motor deficit of HE. The present study aims to investigate the effects of oxidative stress on HE in male mice. The results validate the existence of oxidative stress in both liver and SNr across two murine models of HE induced by thioacetamide (TAA) and bile duct ligation (BDL). Systemic mitochondria-targeted antioxidative drug mitoquinone (Mito-Q) rescues mitochondrial dysfunction and oxidative injury in SNr, so as to restore the locomotor impairment in TAA and BDL mice. Furthermore, the GAD2-expressing SNr population (SNrGAD2) is activated by HE. Both overexpression of mitochondrial uncoupling protein 2 (UCP2) targeted to SNrGAD2 and SNrGAD2-targeted chemogenetic inhibition targeted to SNrGAD2 rescue mitochondrial dysfunction in TAA-induced HE. These results define the key role of oxidative stress in the pathogenesis of HE.

Recent Advances in Sensor Technology for Healthcare and Biomedical Applications (Volume II).

With remarkable progress being witnessed in recent years in the development of sensors, these advances in sensor technology provide unprecedented opportunities for (1) the early diagnosis and prevention of human diseases by detecting critical biomarkers; (2) health assessments by monitoring and analyzing human physiological signals in healthcare and biomedical applications; and (3) the efficient evaluation of human-health-relevant environmental factors by monitoring and measuring environmental determinants [...].

Appropriateness of Empirical Antibiotic Therapy in Hospitalized Patients with Bacterial Infection: A Retrospective Cohort Study.

Objective: The incidence of inappropriate and excessive empirical antibiotic therapy is unclear. The aim of this study was to determine the prevalence of different empirical antibiotic therapy prescriptions, related factors, and outcomes in hospitalized patients with bacterial infection. Methods: A retrospective cohort study was performed and patients with bacterial infection who were admitted between October 1, 2019, and September 30, 2020, were included. Multivariable analysis was performed by the logistic regression model. Results: A total of 536 (42.6%) of the 1257 included patients received inappropriate empirical antibiotic therapy (IEAT), and 368 (29.3%) patients received appropriate but unnecessarily broad-spectrum empirical antibiotic therapy (AUEAT). MDRO (adjusted OR 2.932 [95% CI 2.201~3.905]; p < 0.001) and fever on admission (adjusted OR 0.592 [95% CI 0.415~0.844]; p = 0.004) were correlates of IEAT; sepsis (adjusted OR 2.342 [95% CI 1.371~3.999]; p = 0.002), age (adjusted OR 1.019 [95% CI 1.008~1.030]; p < 0.001), MDRO (adjusted OR 0.664 [95% CI 0.469~0.941]; p = 0.021), and urinary tract infection (adjusted OR 0.352 [95% CI 0.203~0.611]; p < 0.001) were correlates of AUEAT. Patients who received AUEAT were more likely to have a poor prognosis (63 [17.8%] vs 101 [27.4%]; p = 0.002). Both IEAT (median [IQR], 24,971 [13,135-70,155] vs 31,489 [14,894-101,082] CNY; p = 0.007) and AUEAT (median [IQR], 24,971 [13,135-70,155] vs 30,960 [16,475-90,881] CNY; p = 0.002) increased hospital costs. 45.3% (570/1257) of patients were infected with MDRO and 62.9% of them received IEAT. Conclusion: Inappropriate and excessive empirical antibiotic use was widely prevalent among hospitalized patients. Either inappropriate or excessive use of antibiotics may increase the burden of healthcare costs, the latter of which may be associated with poor prognosis. Clinicians need to be more judicious in choosing antibiotic(s). The MDRO epidemic was severe, especially in patients who received IEAT. It is imperative to take effective measures to improve the current situation of antibiotic abuse and antimicrobial resistance.

The Colonization of Rumen Microbiota and Intervention in Pre-Weaned Ruminants.

In pre-weaned ruminants, the microbiota colonizes rapidly in the rumen after birth and constantly interacts with the host to sustain health and metabolism. The developing microbial community is more malleable, so its manipulation may improve ruminant health and productivity as well as may have long-term effects on ruminants. Hence, understanding the process of rumen microbiota establishment is helpful for nutritional interventions of rumen microbiota in pre-weaned ruminants. This paper reviews the latest advances in the colonization of rumen microbiota while providing insights into the most suitable time for manipulating rumen microbial colonization in early life. In addition, different factors that affect rumen microbiota establishment during the pre-weaned ruminants are discussed in the current manuscript. The purpose of this review is to aid in the development of guidelines for manipulating rumen microbiota to improve animal productivity and health.

High-Speed Dicing of SiC Wafers with 0.048 mm Diamond Blades via Rolling-Slitting.

In this study, an innovative fabrication method called rolling-slitting forming, which forms ultra-thin diamond blades, was presented for the first time. Furthermore, the feasibility of the rolling-slitting forming method when applied to silicon carbide wafer dicing blades was investigated; moreover, the cold-pressing blade samples were manufactured through the conventional process under the same sintering conditions to compare and analyze the manufacturing efficiency, organization and performance. The results show that the new method achieves high-precision and low-thickness dicing blades through continuous production without molds-with the thinnest blades being 0.048 mm thick. Furthermore, the rolling-slitting blade has a unique multiporous heat-conductive matrix structure and in-situ generated amorphous pyrolytic carbon, which can reduce the dicing resistance and contribute to a better cutting quality. In addition, the effects of the dicing parameters on SiC were investigated by using indications of spindle current, dicing chipping size and kerf width during the high dicing process. For a dicing depth of 0.2 mm, the ideal performance of dicing SiC with an ultra-thin blade was achieved at a spindle speed of 22,000 rpm and a feed rate of 5 mm/s. This research provides a new idea for the manufacturing of dicing blades, which can satisfy the demand for ultra-narrow dicing streets of high integration of ICs.

Novel fabrication tools for dynamic compression targets with engineered voids using photolithography methods.

Mesoscale imperfections, such as pores and voids, can strongly modify the properties and the mechanical response of materials under extreme conditions. Tracking the material response and microstructure evolution during void collapse is crucial for understanding its performance. In particular, imperfections in the ablator materials, such as voids, can limit the efficiency of the fusion reaction and ultimately hinder ignition. To characterize how voids influence the response of materials during dynamic loading and seed hydrodynamic instabilities, we have developed a tailored fabrication procedure for designer targets with voids at specific locations. Our procedure uses SU-8 as a proxy for the ablator materials and hollow silica microspheres as a proxy for voids and pores. By using photolithography to design the targets' geometry, we demonstrate precise and highly reproducible placement of a single void within the sample, which is key for a detailed understanding of its behavior under shock compression. This fabrication technique will benefit high-repetition rate experiments at x-ray and laser facilities. Insight from shock compression experiments will provide benchmarks for the next generation of microphysics modeling.

Multi-frame, ultrafast, x-ray microscope for imaging shockwave dynamics.

Inertial confinement fusion (ICF) holds increasing promise as a potential source of abundant, clean energy, but has been impeded by defects such as micro-voids in the ablator layer of the fuel capsules. It is critical to understand how these micro-voids interact with the laser-driven shock waves that compress the fuel pellet. At the Matter in Extreme Conditions (MEC) instrument at the Linac Coherent Light Source (LCLS), we utilized an x-ray pulse train with ns separation, an x-ray microscope, and an ultrafast x-ray imaging (UXI) detector to image shock wave interactions with micro-voids. To minimize the high- and low-frequency variations of the captured images, we incorporated principal component analysis (PCA) and image alignment for flat-field correction. After applying these techniques we generated phase and attenuation maps from a 2D hydrodynamic radiation code (xRAGE), which were used to simulate XPCI images that we qualitatively compare with experimental images, providing a one-to-one comparison for benchmarking material performance. Moreover, we implement a transport-of-intensity (TIE) based method to obtain the average projected mass density (areal density) of our experimental images, yielding insight into how defect-bearing ablator materials alter microstructural feature evolution, material compression, and shock wave propagation on ICF-relevant time scales.

Independent associations of short- and long-term air pollution exposure with COVID-19 mortality among Californians.

The growing literature demonstrating air pollution associations on COVID-19 mortality contains studies predominantly examining long-term exposure, with few on short-term exposure, and rarely both together to estimate independent associations. Because mechanisms by which air pollution may impact COVID-19 mortality risk function over timescales ranging from years to days, and given correlation among exposure time windows, consideration of both short- and long-term exposure is of importance. We assessed the independent associations between COVID-19 mortality rates with short- and long-term air pollution exposure by modeling both concurrently. Using California death certificate data COVID-19-related deaths were identified, and decedent residential information used to assess short- (4-week mean) and long-term (6-year mean) exposure to particulate matter <2.5µm (PM2.5), nitrogen dioxide (NO2), and ozone (O3). Negative binomial mixed models were fitted on weekly census tract COVID-19 mortality adjusting for potential confounders with random effects for county and census tract and an offset for population. Data were evaluated separately for two time periods March 16, 2020-October 18, 2020 and October 19, 2020-April 25, 2021, representing the Spring/Summer surges and Winter surge. Independent positive associations with COVID-19 mortality were observed for short- and long-term PM2.5 in both study periods, with strongest associations observed in the first study period: COVID-19 mortality rate ratio for a 2-µg/m3 increase in long-term PM2.5 was 1.13 (95%CI:1.09,1.17) and for a 4.7-µg/m3 increase in short-term PM2.5 was 1.05 (95%CI:1.02,1.08). Statistically significant positive associations were seen for both short- and long-term NO2 in study period 1, but short-term NO2 was not statistically significant in study period 2. Results for long-term O3 indicate positive associations, however, only marginal significance is achieved in study period 1. These findings support an adverse effect of long-term PM2.5 and NO2 exposure on COVID-19 mortality risk, independent of short-term exposure, and a possible independent effect of short-term PM2.5.

AIF Overexpression Aggravates Oxidative Stress in Neonatal Male Mice After Hypoxia-Ischemia Injury.

There are sex differences in the severity, mechanisms, and outcomes of neonatal hypoxia-ischemia (HI) brain injury, and apoptosis-inducing factor (AIF) may play a critical role in this discrepancy. Based on previous findings that AIF overexpression aggravates neonatal HI brain injury, we further investigated potential sex differences in the severity and molecular mechanisms underlying the injury using mice that overexpress AIF from homozygous transgenes. We found that the male sex significantly aggravated AIF-driven brain damage, as indicated by the injury volume in the gray matter (2.25 times greater in males) and by the lost volume of subcortical white matter (1.71 greater in males) after HI. As compared to females, male mice exhibited more severe brain injury, correlating with reduced antioxidant capacities, more pronounced protein carbonylation and nitration, and increased neuronal cell death. Under physiological conditions (without HI), the doublecortin-positive area in the dentate gyrus of females was 1.15 times larger than in males, indicating that AIF upregulation effectively promoted neurogenesis in females in the long term. We also found that AIF stimulated carbohydrate metabolism in young males. Altogether, these findings corroborate earlier studies and further demonstrate that AIF is involved in oxidative stress, which contributes to the sex-specific differences observed in neonatal HI brain injury.

Dual inhibition of autophagy and PI3K/mTOR pathway as a potential therapeutic strategy against laryngeal squamous cell carcinoma.

Background: New and effective chemotherapy or targeted therapy strategies are needed against laryngeal squamous cell carcinoma (LSCC). We aimed to explore the antitumor effect of dual PI3K/mTOR inhibitor combined with autophagy suppression on LSCC and its underlying mechanism. Methods: Hep-2 and AMC-HN-8 cell lines were treated with the Akt inhibitor LY294002, mTOR inhibitor rapamycin, and dual inhibitor NVP-BEZ235 separately. The biological characteristics of in vitro proliferation, cell cycle, apoptosis, migration, invasion, and autophagy were analyzed, and the expression levels of PI3K/Akt/mTOR pathway-related proteins were also measured. The in vivo effects of NVP-BEZ235 combined with inhibition of autophagy using pharmacological inhibitor was further assessed. Results: Compared with Akt or mTOR inhibitor, NVP-BEZ235 had the most significant biological effects on LSCC cells. When combined with various autophagy inhibitors, along with siRNA against ATG7, NVP-BEZ235 showed a synergic antitumor effect in LSCC through increasing cell apoptosis and death both in vitro and vivo. Conclusions: NVP-BEZ235 exerted potent antitumor effects on LSCC, especially when combined with the autophagy inhibitor both in vitro and vivo, providing convincing experimental data for new molecular targeted therapy for LSCC.

Long-term air pollution and COVID-19 mortality rates in California: Findings from the Spring/Summer and Winter surges of COVID-19.

A growing number of studies report associations between air pollution and COVID-19 mortality. Most were ecological studies at the county or regional level which disregard important local variability and relied on data from only the first few months of the pandemic. Using COVID-19 deaths identified from death certificates in California, we evaluated whether long-term ambient air pollution was related to weekly COVID-19 mortality at the census tract-level during the first ∼12 months of the pandemic. Weekly COVID-19 mortality for each census tract was calculated based on geocoded death certificate data. Annual average concentrations of ambient particulate matter <2.5 µm (PM2.5) and <10 µm (PM10), nitrogen dioxide (NO2), and ozone (O3) over 2014-2019 were assessed for all census tracts using inverse distance-squared weighting based on data from the ambient air quality monitoring system. Negative binomial mixed models related weekly census tract COVID-19 mortality counts to a natural cubic spline for calendar week. We included adjustments for potential confounders (census tract demographic and socioeconomic factors), random effects for census tract and county, and an offset for census tract population. Data were analyzed as two study periods: Spring/Summer (March 16-October 18, 2020) and Winter (October 19, 2020-March 7, 2021). Mean (standard deviation) concentrations were 10.3 (2.1) µg/m3 for PM2.5, 25.5 (7.1) µg/m3 for PM10, 11.3 (4.0) ppb for NO2, and 42.8 (6.9) ppb for O3. For Spring/Summer, adjusted rate ratios per standard deviation increase were 1.13 (95% confidence interval: 1.09, 1.17) for PM2.5, 1.16 (1.11, 1.21) for PM10, 1.06 (1.02, 1.10) for NO2, and 1.09 (1.04, 1.14) for O3. Associations were replicated in Winter, although they were attenuated for PM2.5 and PM10. Study findings support a relation between long-term ambient air pollution exposure and COVID-19 mortality. Communities with historically high pollution levels might be at higher risk of COVID-19 mortality.

Using dynamic time warping self-organizing maps to characterize diurnal patterns in environmental exposures.

Advances in measurement technology are producing increasingly time-resolved environmental exposure data. We aim to gain new insights into exposures and their potential health impacts by moving beyond simple summary statistics (e.g., means, maxima) to characterize more detailed features of high-frequency time series data. This study proposes a novel variant of the Self-Organizing Map (SOM) algorithm called Dynamic Time Warping Self-Organizing Map (DTW-SOM) for unsupervised pattern discovery in time series. This algorithm uses DTW, a similarity measure that optimally aligns interior patterns of sequential data, both as the similarity measure and training guide of the neural network. We applied DTW-SOM to a panel study monitoring indoor and outdoor residential temperature and particulate matter air pollution (PM2.5) for 10 patients with asthma from 7 households near Salt Lake City, UT; the patients were followed for up to 373 days each. Compared to previous SOM algorithms using timestamp alignment on time series data, the DTW-SOM algorithm produced fewer quantization errors and more detailed diurnal patterns. DTW-SOM identified the expected typical diurnal patterns in outdoor temperature which varied by season, as well diurnal patterns in PM2.5 which may be related to daily asthma outcomes. In summary, DTW-SOM is an innovative feature engineering method that can be applied to highly time-resolved environmental exposures assessed by sensors to identify typical diurnal (or hourly or monthly) patterns and provide new insights into the health effects of environmental exposures.

Structural Basis of a Human Neutralizing Antibody Specific to the SARS-CoV-2 Spike Protein Receptor-Binding Domain.

The emerging new lineages of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) have marked a new phase of coronavirus disease 2019 (COVID-19). Understanding the recognition mechanisms of potent neutralizing monoclonal antibodies (NAbs) against the spike protein is pivotal for developing new vaccines and antibody drugs. Here, we isolated several monoclonal antibodies (MAbs) against the SARS-CoV-2 spike protein receptor-binding domain (S-RBD) from the B cell receptor repertoires of a SARS-CoV-2 convalescent. Among these MAbs, the antibody nCoV617 demonstrates the most potent neutralizing activity against authentic SARS-CoV-2 infection, as well as prophylactic and therapeutic efficacies against the human angiotensin-converting enzyme 2 (ACE2) transgenic mouse model in vivo. The crystal structure of S-RBD in complex with nCoV617 reveals that nCoV617 mainly binds to the back of the "ridge" of RBD and shares limited binding residues with ACE2. Under the background of the S-trimer model, it potentially binds to both "up" and "down" conformations of S-RBD. In vitro mutagenesis assays show that mutant residues found in the emerging new lineage B.1.1.7 of SARS-CoV-2 do not affect nCoV617 binding to the S-RBD. These results provide a new human-sourced neutralizing antibody against the S-RBD and assist vaccine development. IMPORTANCE COVID-19 is a respiratory disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The COVID-19 pandemic has posed a serious threat to global health and the economy, so it is necessary to find safe and effective antibody drugs and treatments. The receptor-binding domain (RBD) in the SARS-CoV-2 spike protein is responsible for binding to the angiotensin-converting enzyme 2 (ACE2) receptor. It contains a variety of dominant neutralizing epitopes and is an important antigen for the development of new coronavirus antibodies. The significance of our research lies in the determination of new epitopes, the discovery of antibodies against RBD, and the evaluation of the antibodies' neutralizing effect. The identified antibodies here may be drug candidates for the development of clinical interventions for SARS-CoV-2.

W-TSS: A Wavelet-Based Algorithm for Discovering Time Series Shapelets.

Many approaches to time series classification rely on machine learning methods. However, there is growing interest in going beyond black box prediction models to understand discriminatory features of the time series and their associations with outcomes. One promising method is time-series shapelets (TSS), which identifies maximally discriminative subsequences of time series. For example, in environmental health applications TSS could be used to identify short-term patterns in exposure time series (shapelets) associated with adverse health outcomes. Identification of candidate shapelets in TSS is computationally intensive. The original TSS algorithm used exhaustive search. Subsequent algorithms introduced efficiencies by trimming/aggregating the set of candidates or training candidates from initialized values, but these approaches have limitations. In this paper, we introduce Wavelet-TSS (W-TSS) a novel intelligent method for identifying candidate shapelets in TSS using wavelet transformation discovery. We tested W-TSS on two datasets: (1) a synthetic example used in previous TSS studies and (2) a panel study relating exposures from residential air pollution sensors to symptoms in participants with asthma. Compared to previous TSS algorithms, W-TSS was more computationally efficient, more accurate, and was able to discover more discriminative shapelets. W-TSS does not require pre-specification of shapelet length.