Showing papers by "National University of Singapore published in 2023"

Unintended consequences of lockdowns, COVID-19 and the Shadow Pandemic in India

Abstract: Violence against women is a problem worldwide, with economic costs ranging from 1% to 4% of global gross domestic product. During the coronavirus disease 2019 lockdowns, the United Nations coined the term the Shadow Pandemic to describe the increase in global violence against women. Here, using variation in the intensity of government-mandated lockdowns in India, we show that domestic violence complaints increase significantly in districts with the strictest lockdown rules. We find similarly large increases in cybercrime complaints. However, rape and sexual assault complaints decrease in districts with the strictest lockdowns, consistent with decreased female mobility in public spaces, public transport and workplaces where they might be at greater risk for rape and sexual assault. Medium-term analysis shows that increases in domestic violence complaints persist 1 year later, while other complaints related to rape, sexual assault and cybercrimes return to pre-lockdown levels. Lockdowns may help control disease, but also come with potential costs. Domestic violence complaints in India increased in districts with the strictest lockdown rules, and remained higher 1 year later, even after restrictions were loosened.

Physics-Informed Online Machine Learning and Predictive Control of Nonlinear Processes with Parameter Uncertainty

Abstract: In this work, we present a physics-informed recurrent neural network (PIRNN)-based modeling approach for nonlinear dynamic systems with parameter uncertainty. Physics-informed modeling approaches can improve the generalization performance of machine learning models by embedding the knowledge of physical laws in the learning process. Based on the standard PIRNN modeling approach for the nominal system without model uncertainties, we develop a novel PIRNN-enhanced modeling method that integrates online estimation of uncertain process parameters into the training process using the latest process data. The PIRNN-enhanced modeling approach is incorporated into the design of model predictive control (MPC), where an error-triggered mechanism is employed to trigger the quantification of modeling uncertainties and update of PIRNN models accordingly. Finally, a chemical process example is used to demonstrate the superiority of the proposed PIRNN-enhanced online learning mechanism in comparison to the conventional purely data-driven online update strategy for nonlinear systems subject to process parameter uncertainty.

Actin polymerisation and crosslinking drive left-right asymmetry in single cell and cell collectives

Abstract: Deviations from mirror symmetry in the development of bilateral organisms are common but the mechanisms of initial symmetry breaking are insufficiently understood. The actin cytoskeleton of individual cells self-organises in a chiral manner, but the molecular players involved remain essentially unidentified and the relationship between chirality of an individual cell and cell collectives is unclear. Here, we analysed self-organisation of the chiral actin cytoskeleton in individual cells on circular or elliptical patterns, and collective cell alignment in confined microcultures. Screening based on deep-learning analysis of actin patterns identified actin polymerisation regulators, depletion of which suppresses chirality (mDia1) or reverses chirality direction (profilin1 and CapZβ). The reversed chirality is mDia1-independent but requires the function of actin-crosslinker α-actinin1. A robust correlation between the effects of a variety of actin assembly regulators on chirality of individual cells and cell collectives is revealed. Thus, actin-driven cell chirality may underlie tissue and organ asymmetry.

Grammatical Error Correction: A Survey of the State of the Art

Abstract: Abstract Grammatical Error Correction (GEC) is the task of automatically detecting and correcting errors in text. The task not only includes the correction of grammatical errors, such as missing prepositions and mismatched subject–verb agreement, but also orthographic and semantic errors, such as misspellings and word choice errors, respectively. The field has seen significant progress in the last decade, motivated in part by a series of five shared tasks, which drove the development of rule-based methods, statistical classifiers, statistical machine translation, and finally neural machine translation systems, which represent the current dominant state of the art. In this survey paper, we condense the field into a single article and first outline some of the linguistic challenges of the task, introduce the most popular datasets that are available to researchers (for both English and other languages), and summarize the various methods and techniques that have been developed with a particular focus on artificial error generation. We next describe the many different approaches to evaluation as well as concerns surrounding metric reliability, especially in relation to subjective human judgments, before concluding with an overview of recent progress and suggestions for future work and remaining challenges.We hope that this survey will serve as comprehensive resource for researchers who are new to the field or who want to be kept apprised of recent developments.

Early morning university classes are associated with impaired sleep and academic performance

Abstract: Abstract Attending classes and sleeping well are important for students’ academic success. Here, we tested whether early morning classes are associated with lower attendance, shorter sleep and poorer academic achievement by analysing university students’ digital traces. Wi-Fi connection logs in 23,391 students revealed that lecture attendance was about ten percentage points lower for classes at 08:00 compared with later start times. Diurnal patterns of Learning Management System logins in 39,458 students and actigraphy data in 181 students demonstrated that nocturnal sleep was an hour shorter for early classes because students woke up earlier than usual. Analyses of grades in 33,818 students showed that the number of days per week they had morning classes was negatively correlated with grade point average. These findings suggest concerning associations between early morning classes and learning outcomes.

Facile Synthesis and Chiral Resolution of Expanded Helicenes with up to 35 <i>cata</i> ‐Fused Benzene Rings**

Abstract: Expanded helicenes are expected to show enhanced chiroptical properties as compared to the classical helicenes but the synthesis is very challenging. Herein, we report the facile synthesis of a series of expanded helicenes Hn (n=1-4) containing 11, 19, 27 and 35 cata-fused benzene rings through Suzuki coupling-based oligomerization followed by Bi(OTf)3 -mediated regioselective cyclization of vinyl ethers. Their structures were determined by X-ray crystallographic analysis. Enantiopure H2, H3, and H4 can be isolated by chiral HPLC and they all exhibit strong chiroptical responses with high absorption dissymmetry factor (|gabs |) values (0.020 for H2, 0.021 for H3, and 0.021-0.024 for H4).

Interpenetrating phase composites with 3D printed triply periodic minimal surface (TPMS) lattice structures

Abstract: Micro-selective laser melting fabricated triply periodic minimal surface sheet lattice and epoxy interpenetrating phase composites (IPCs) are investigated in this paper. To achieve superior specific energy absorption (SEA) enhancements in the composites, an optimised Schwarz primitive lattice (P-lattice) structure is proposed by redefining the shell opening diameter with a shape parameter. Also, the influences of fabrication directions, along the [100], [110], and [111] lattice directions, on the mechanical responses of the P-lattices and IPCs are presented. Compression results reveal that the modified P-lattices outperform the original P-lattices with superior compressive strength and SEA. The P-lattices also display the lowest strength and SEA along [100] as compared to that of [110] and [111]. The compressive strength and SEA of modified P-lattice along [111] are 123.66% and 64.63% higher than its [100], respectively. As for the IPCs, up to a 52% increase, from the linear addition of the two-component phases has been achieved for SEA. The IPCs also exhibit a superior specific energy absorption of 49.6 J/g, a 1109% improvement from that of the pure lattice, which is attributed to the high strength and large plateau strain of the composites. The simulation results show that the internal energy of both lattice and epoxy in composites is 136% and 21%, respectively, higher than that of single structures due to the interaction effects.

Cooling photoluminescent phosphors in laser-excited white lighting with three-dimensional boron nitride networks

Abstract: Laser-excited white lighting, which is achieved by laser diodes and phosphor converters, has garnered increasing interest in recent years due to their high-power and high-brightness characteristics. Nevertheless, two long-standing thermal issues remain formidable: first, the photoluminescent phosphors also generate heat due to the Stokes loss, which are the second heat source in the packaging but long ignored; second, the phosphors are embedded in low-thermal-conductivity silicone and the generated heat is rather inefficient to dissipate. As a result, the accumulated heat leads to high temperature in phosphor particles and even thermal quenching, which deteriorate the luminescence, chromaticity, and reliability. In this work, these problems are successfully addressed by establishing three-dimensional (3D) boron nitride (BN) networks to cool the silicone-embedded phosphor particles in a facile and scalable method. When lighted under the driven current of 900 mA, the working temperature of phosphor layer is dramatically decreased by 95.2 °C with the 3D-BN network without sacrificing the luminescence and chromaticity. Such strategy with excellent cooling performance, high reliability, and easy scalability may hold great promise for broader optoelectronics applications beyond laser-excited white lighting.

Area-Under-Curve–Guided Versus Trough-Guided Monitoring of Vancomycin and Its Impact on Nephrotoxicity: A Systematic Review and Meta-Analysis

Abstract: Conventionally, vancomycin trough levels have been used for therapeutic drug monitoring (TDM). Owing to the increasing evidence of trough levels being poor surrogates of area under the curve (AUC) and the advent of advanced pharmacokinetics software, a paradigm shift has been made toward AUC-guided dosing. This study aims to evaluate the impact of AUC-guided versus trough-guided TDM on vancomycin-associated nephrotoxicity.A systematic review was conducted using PubMed, Embase, Web of Science, Cumulative Index to Nursing and Allied Health Literature, Google scholar, and Cochrane library databases; articles published from January 01, 2009, to January 01, 2021, were retrieved and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses checklist. Studies that evaluated trough-guided or AUC-guided vancomycin TDM and vancomycin-associated nephrotoxicity were included. Random-effects models were used to compare the differences in nephrotoxicity.Of the 1191 retrieved studies, 57 were included. Most studies included adults and older adults (n = 47, 82.45%). The pooled prevalence of nephrotoxicity was lower in AUC-guided TDM [6.2%; 95% confidence interval (CI): 2.9%-9.5%] than in trough-guided TDM (17.0%; 95% CI: 14.7%-19.2%). Compared with the trough-guided approach, the AUC-guided approach had a lower risk of nephrotoxicity (odds ratio: 0.53; 95% CI: 0.32-0.89). The risk of nephrotoxicity was unaffected by the AUC derivation method. AUC thresholds correlated with nephrotoxicity only within the first 96 hours of therapy.The AUC-guided approach had a lower risk of nephrotoxicity, supporting the updated American Society of Health-System Pharmacists guidelines. Further studies are needed to evaluate the optimal AUC-derivation methods and clinical utility of repeated measurements of the AUC and trough levels of vancomycin.

A Machine Learning Approach for Early Diagnosis of Cognitive Impairment Using Population-Based Data

Abstract: The major mechanisms of dementia and cognitive impairment are vascular and neurodegenerative processes. Early diagnosis of cognitive impairment can facilitate timely interventions to mitigate progression.This study aims to develop a reliable machine learning (ML) model using socio-demographics, vascular risk factors, and structural neuroimaging markers for early diagnosis of cognitive impairment in a multi-ethnic Asian population.The study consisted of 911 participants from the Epidemiology of Dementia in Singapore study (aged 60- 88 years, 49.6% male). Three ML classifiers, logistic regression, support vector machine, and gradient boosting machine, were developed. Prediction results of independent classifiers were combined in a final ensemble model. Model performances were evaluated on test data using F1 score and area under the receiver operating curve (AUC) methods. Post modelling, SHapely Additive exPlanation (SHAP) was applied on the prediction results to identify the predictors that contribute most to the cognitive impairment prediction.The final ensemble model achieved a F1 score and AUC of 0.87 and 0.80 respectively. Accuracy (0.83), sensitivity (0.86), specificity (0.74) and predictive values (positive 0.88 negative 0.72) of the ensemble model were higher compared to the independent classifiers. Age, ethnicity, highest education attainment and neuroimaging markers were identified as important predictors of cognitive impairment.This study demonstrates the feasibility of using ML tools to integrate multiple domains of data for reliable diagnosis of early cognitive impairment. The ML model uses easy-to-obtain variables and is scalable for screening individuals with a high risk of developing dementia in a population-based setting.

Repurposing of Doxycycline to Attenuate Influenza Virus Pathogenesis Via Inhibition of Matrix Metalloproteinases in Neutrophils

Abstract: Influenza viruses constitute a significant threat to public health worldwide over many decades, and continue to inflict significant morbidity and mortality. Previous studies show that infection of mice with mouse-adapted influenza A/Aichi/2/1968(H3N2) passage 10 (P10) virus can elicit exuberant inflammatory responses in the lungs with extensive infiltration of macrophages and neutrophils (which are sources of gelatinases) that contribute to pulmonary damage. The lungs of mice with severe influenza pneumonitis also reveal extensive neutrophilic infiltration, neutrophil extracellular traps (NETs), alveolar damage, heightened viral load, and pathologic features of acute respiratory distress syndrome (ARDS). Excessive neutrophil and matrix metalloproteinase (MMP) activities are implicated in the pathogenesis of acute lung injury (ALI) and ARDS. Hence, an objective of this study was to investigate the production of MMP-2 and MMP-9 in neutrophils differentiated from the MPRO murine pro-myelocytic cell line, following infection with mouse-adapted influenza H3N2 virus. Another objective was to investigate the effects of doxycycline on expression of MMP-2 and MMP-9 at transcriptional and translational levels in neutrophils in the context of influenza virus infection. MMP-2 and MMP-9 production and gelatinase activity were found to be induced by infection of differentiated neutrophils with mouse-adapted influenza H3N2 virus. Furthermore, doxycycline treatment was able to abrogate this increase in MMP-2 and MMP-9 expression and gelatinase activity. Given that excessive neutrophil infiltration and uncontrolled neutrophil gelatinase activity in pulmonary tissues play pivotal roles in the pathogenesis of ALI and ARDS, this study offers insights into the mechanism of excessive MMP-2 and MMP-9 production and activity in neutrophils during influenza virus infection. Thus, repurposing doxycycline as an MMP inhibitor represents a potential therapeutic strategy to ameliorate influenza-associated pulmonary injury by targeting MMPs, gelatinase production and activity in neutrophils. Doxycycline therapy (alone or in combination with other drugs) has also been considered and explored for the management of other infectious diseases such as coronavirus infections and tuberculosis.

Facile Synthesis and Chiral Resolution of Expanded Helicenes with up to 35 <i>cata</i> ‐Fused Benzene Rings**

Abstract: Expanded helicenes are expected to show enhanced chiroptical properties as compared to the classical helicenes but the synthesis is very challenging. Herein, we report the facile synthesis of a series of expanded helicenes Hn (n=1–4) containing 11, 19, 27 and 35 cata-fused benzene rings through Suzuki coupling-based oligomerization followed by Bi(OTf)3-mediated regioselective cyclization of vinyl ethers. Their structures were determined by X-ray crystallographic analysis. Enantiopure H2, H3, and H4 can be isolated by chiral HPLC and they all exhibit strong chiroptical responses with high absorption dissymmetry factor (|gabs|) values (0.020 for H2, 0.021 for H3, and 0.021–0.024 for H4).

Quantum noise spectroscopy as an incoherent imaging problem

Abstract: I point out the mathematical correspondence between an incoherent imaging model proposed by my group in the study of quantum-inspired superresolution [M. Tsang, R. Nair, and X.-M. Lu, Phys. Rev. X 6, 031033 (2016)] and a noise spectroscopy model also proposed by us [M. Tsang and R. Nair, Phys. Rev. A 86, 042115 (2012); S. Ng et al., Phys. Rev. A 93, 042121 (2016)]. Both can be regarded as random displacement models, where the probability measure for the random displacement depends on unknown parameters. The spatial-mode demultiplexing (SPADE) method proposed for imaging is analogous to the spectral photon counting method proposed by Ng et al. for optical phase noise spectroscopy: Both methods are discrete-variable measurements that are superior to direct displacement measurements (direct imaging or homodyne detection) and can achieve the respective quantum limits. Inspired by SPADE, I propose a modification of spectral photon counting when the input field is squeezed: The output field is simply unsqueezed before spectral photon counting. I show that this method is quantum optimal and far superior to homodyne detection for both parameter estimation and detection, thus solving the open problems in the work of Tsang and Nair and Ng et al.

Time-resolved assessment of single-cell protein secretion by sequencing

Abstract: Secreted proteins play critical roles in cellular communication. Methods enabling concurrent measurement of cellular protein secretion, phenotypes and transcriptomes are still unavailable. Here we describe time-resolved assessment of protein secretion from single cells by sequencing (TRAPS-seq). Released proteins are trapped onto the cell surface and probed by oligonucleotide-barcoded antibodies before being simultaneously sequenced with transcriptomes in single cells. We demonstrate that TRAPS-seq helps unravel the phenotypic and transcriptional determinants of the secretion of pleiotropic TH1 cytokines (IFNγ, IL-2 and TNF) in activated T cells. In addition, we show that TRAPS-seq can be used to track the secretion of multiple cytokines over time, uncovering unique molecular signatures that govern the dynamics of single-cell cytokine secretions. Our results revealed that early central memory T cells with CD45RA expression (TCMRA) are important in both the production and maintenance of polyfunctional cytokines. TRAPS-seq presents a unique tool for seamless integration of secretomics measurements with multi-omics profiling in single cells. TRAPS-seq is a sequencing-based method to profile dynamic protein secretion in single cells.

Clinical Progress in Gold Nanoparticle (GNP)-mediated Photothermal Cancer Therapy

Abstract: Abstract: The field of oncotherapy has always been looking out for alternative treatment methods that have much lesser side effects compared to the currently used therapies that lower the patients’ quality of life. Gold Nanoparticle (GNP)-mediated photothermal therapies are proving to be a boon as they are both non-invasive and tumour-specific. This review analyses how GNPs can help right from the beginning, that is, the diagnosis of cancer, to the end, that is, effective ablation of cancerous cells. Their ability to function as photothermal absorbers, targeted drug deliverers, and inducers of photoimmunity are reviewed in detail, bringing out the current clinical progress in each of those areas. Even though they stand to be a promising solution for cancer therapy, it is necessary to understand their biodegradation and in vivo toxicity before their extensive clinical usage.