RESUMO
Network-based machine learning (ML) has the potential for predicting novel genes associated with nearly any health and disease context. However, this approach often uses network information from only the single species under consideration even though networks for most species are noisy and incomplete. While some recent methods have begun addressing this shortcoming by using networks from more than one species, they lack one or more key desirable properties: handling networks from more than two species simultaneously, incorporating many-to-many orthology information, or generating a network representation that is reusable across different types of and newly-defined prediction tasks. Here, we present GenePlexusZoo, a framework that casts molecular networks from multiple species into a single reusable feature space for network-based ML. We demonstrate that this multi-species network representation improves both gene classification within a single species and knowledge-transfer across species, even in cases where the inter-species correspondence is undetectable based on shared orthologous genes. Thus, GenePlexusZoo enables effectively leveraging the high evolutionary molecular, functional, and phenotypic conservation across species to discover novel genes associated with diverse biological contexts.
Assuntos
Genômica , Aprendizado de Máquina , Genômica/métodosRESUMO
Engineering room-temperature strong coupling of few-exciton in transition-metal dichalcogenides (TMDCs) with plasmons promises to construct compact and high-performance quantum optical devices. But it remains unimplemented due to their in-plane excitons. Here, we demonstrate the strong coupling of few-exciton within 10 in monolayer WS2 with the plasmonic mode with a large tangential component of the electric field tightly trapped around the sharp corners of an Au@Ag nanocuboid, the fewest number of excitons observed in the TMDC family so far. Furthermore, we for the first time report a significant deviation with a relative difference of up to 100.6% between the spectrum and eigenlevel splitting dispersions, which increases with decreasing coupling strength. It is also shown that the coupling strength obtained by the conventional concept of both being equal to the measured spectrum splitting is markedly overestimated. Our work enriches the understanding of strong light-matter interactions at room temperature.
RESUMO
MOTIVATION: Accurately representing biological networks in a low-dimensional space, also known as network embedding, is a critical step in network-based machine learning and is carried out widely using node2vec, an unsupervised method based on biased random walks. However, while many networks, including functional gene interaction networks, are dense, weighted graphs, node2vec is fundamentally limited in its ability to use edge weights during the biased random walk generation process, thus under-using all the information in the network. RESULTS: Here, we present node2vec+, a natural extension of node2vec that accounts for edge weights when calculating walk biases and reduces to node2vec in the cases of unweighted graphs or unbiased walks. Using two synthetic datasets, we empirically show that node2vec+ is more robust to additive noise than node2vec in weighted graphs. Then, using genome-scale functional gene networks to solve a wide range of gene function and disease prediction tasks, we demonstrate the superior performance of node2vec+ over node2vec in the case of weighted graphs. Notably, due to the limited amount of training data in the gene classification tasks, graph neural networks such as GCN and GraphSAGE are outperformed by both node2vec and node2vec+. AVAILABILITY AND IMPLEMENTATION: The data and code are available on GitHub at https://github.com/krishnanlab/node2vecplus_benchmarks. All additional data underlying this article are available on Zenodo at https://doi.org/10.5281/zenodo.7007164. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
Assuntos
Aprendizado de Máquina , Redes Neurais de Computação , Redes Reguladoras de Genes , Fenótipo , Epistasia GenéticaRESUMO
SUMMARY: PyGenePlexus is a Python package that enables a user to gain insight into any gene set of interest through a molecular interaction network informed supervised machine learning model. PyGenePlexus provides predictions of how associated every gene in the network is to the input gene set, offers interpretability by comparing the model trained on the input gene set to models trained on thousands of known gene sets, and returns the network connectivity of the top predicted genes. AVAILABILITY AND IMPLEMENTATION: https://pypi.org/project/geneplexus/ and https://github.com/krishnanlab/PyGenePlexus. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
Assuntos
Biologia Computacional , Software , Aprendizado de Máquina , Aprendizado de Máquina Supervisionado , Estudos de Associação GenéticaRESUMO
Biomedical researchers take advantage of high-throughput, high-coverage technologies to routinely generate sets of genes of interest across a wide range of biological conditions. Although these technologies have directly shed light on the molecular underpinnings of various biological processes and diseases, the list of genes from any individual experiment is often noisy and incomplete. Additionally, interpreting these lists of genes can be challenging in terms of how they are related to each other and to other genes in the genome. In this work, we present GenePlexus (https://www.geneplexus.net/), a web-server that allows a researcher to utilize a powerful, network-based machine learning method to gain insights into their gene set of interest and additional functionally similar genes. Once a user uploads their own set of human genes and chooses between a number of different human network representations, GenePlexus provides predictions of how associated every gene in the network is to the input set. The web-server also provides interpretability through network visualization and comparison to other machine learning models trained on thousands of known process/pathway and disease gene sets. GenePlexus is free and open to all users without the need for registration.
Assuntos
Computadores , Software , Humanos , Genoma , Aprendizado de Máquina , Estudos de Associação Genética , InternetRESUMO
The single-exciton strong coupling with the localized plasmon mode (LPM) at room temperature is highly desirable for exploiting quantum technology. However, its realization has been a very low probability event due to the harsh critical conditions, severely compromising its application. Here, we present a highly efficient approach for achieving such a strong coupling by reducing the critical interaction strength at the exceptional point based upon the damping inhibition and matching of the coupled system, instead of enhancing the coupling strength to overcome the system's large damping. Experimentally, we compress the LPM's damping linewidth from about 45 nm to about 14 nm using a leaky Fabry-Perot cavity, a good match to the excitonic linewidth of about 10 nm. This method dramatically relaxes the harsh requirement in mode volume by more than an order of magnitude and allows a maximum direction angle of the exciton dipole relative to the mode field of up to around 71.9°, significantly improving the success rate of achieving the single-exciton strong coupling with LPMs from about 1% to about 80%.
RESUMO
AIMS: Metabolic syndrome (MetS) is associated with arrhythmias and cardiovascular mortality. Arrhythmogenesis in MetS results from atrial structural and electrical remodelling. The small-conductance Ca2+-activated K+ (SK) currents modulate atrial repolarization and may influence atrial arrhythmogenicity. This study investigated the regulation of SK current perturbed by a high-fat diet (HFD) to mimic MetS. METHODS AND RESULTS: Thirty mice were divided into two groups that were fed with normal chow (CTL) and HFD for 4 months. Electrocardiography and echocardiography were used to detect cardiac electrical and structure remodelling. Atrial action potential duration (APD) and calcium transient duration (CaTD) were measured by optical mapping of Langendorff-perfused mice hearts. Atrial fibrillation (AF) inducibility and duration were assessed by burst pacing. Whole-cell patch clamp was performed in primarily isolated atrial myocytes for SK current density. The SK current density is higher in atrial myocytes from HFD than in CTL mice (P ≤ 0.037). The RNA and protein expression of SK channels are increased in HFD mice (P ≤ 0.041 and P ≤ 0.011, respectively). Action potential duration is shortened in HFD compared with CTL (P ≤ 0.015). The shortening of the atrial APD in HFD is reversed by the application of 100â nM apamin (P ≤ 0.043). Compared with CTL, CaTD is greater in HFD atria (P ≤ 0.029). Calcium transient decay (Tau) is significantly higher in HFD than in CTL (P = 0.001). Both APD and CaTD alternans thresholds were higher in HFD (P ≤ 0.043), along with higher inducibility and longer duration of AF in HFD (P ≤ 0.023). CONCLUSION: Up-regulation of apamin-sensitive SK currents plays a partial role in the atrial arrhythmogenicity of HFD mice.
Assuntos
Fibrilação Atrial , Cálcio , Camundongos , Animais , Cálcio/metabolismo , Potássio/metabolismo , Apamina/metabolismo , Canais de Potássio Ativados por Cálcio de Condutância Baixa/genética , Potenciais de Ação/fisiologia , Miócitos Cardíacos/metabolismoRESUMO
A single quantum dot (QD) strongly coupled with a plasmonic nanoparticle yields a promising qubit for scalable solid-state quantum information processing at room temperature. However, realizing such a strong coupling remains challenging due to the difficulty of spatial overlap of the QD excitons with the plasmonic electric fields (EFs). Here, by using a transmission electron microscope we demonstrate for the first time that this overlap can be realized by integrating a deterministic single QD with a single Au nanorod. When a wedge nanogap cavity consisting of them and the substrate is constructed, the plasmonic EFs can be more effectively "dragged" and highly confined in the QD's nanoshell where the excitons mainly reside. With these advantages, we observed the largest spectral Rabi splitting (reported so far) of â¼234 meV for a single QD strong coupling with plasmons. Our work opens a pathway to the massive construction of room-temperature strong coupling solid qubits.
RESUMO
SUMMARY: Learning low-dimensional representations (embeddings) of nodes in large graphs is key to applying machine learning on massive biological networks. Node2vec is the most widely used method for node embedding. However, its original Python and C++ implementations scale poorly with network density, failing for dense biological networks with hundreds of millions of edges. We have developed PecanPy, a new Python implementation of node2vec that uses cache-optimized compact graph data structures and precomputing/parallelization to result in fast, high-quality node embeddings for biological networks of all sizes and densities. AVAILABILITYAND IMPLEMENTATION: PecanPy software is freely available at https://github.com/krishnanlab/PecanPy. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
RESUMO
An intense terahertz laser field is shown to actively manipulate the electronic states, as well as the linear and nonlinear optical absorption coefficients, of the laterally-coupled quantum well wires (LCQWWs). The laser-dressed quantum states of the LCQWWs are achieved using the non-perturbative Floquet method and the two-dimensional diagonalization technique under the effective mass approximation. We have demonstrated that the intense terahertz laser field induces a strong deformation of the confinement potential configuration of the LCQWWs, thus pronouncedly dressing the energy levels and wave functions. An unambiguous picture is depicted for the evolution of the laser-dressed quantum states with the increase of the laser-dressed parameter characterizing the strength of the laser-dressed effect. On this basis, the resonant peak positions of the linear and nonlinear optical absorption coefficients feature a blue shift followed by a red shift with an increase of the laser-dressed parameter. Furthermore, the evolution of the peak values for the linear and third-order nonlinear optical absorption coefficients as a function of the laser-dressed parameter is comprehensively discussed. Moreover, in contrast to the case without intense terahertz laser field, the peak values of the linear, third-order nonlinear, and total optical absorption coefficients can be obviously enhanced at the same frequency position by manipulating the appropriate laser-dressed parameter. A similar feature can be found in the linear, third-order nonlinear, and total refractive index changes. Our findings are conducive to the implementation of the expected quantum states and nonlinear optical effects in the LCQWWs, paving the way for new designs in tunable optical switches, infrared photo-detectors and infrared modulators.
RESUMO
Here, we demonstrate the first example of 3-isothiocyanato thiobutyrolactone serving as a useful building block in the Michael/cyclization reaction with alkylidene pyrazolones for the enantioselective construction of optically active structural bispiro[pyrazolone-thiobutyrolactone] skeletons containing three contiguous stereocenters with two spiroquaternary stereocenters. These products were smoothly afforded in up to 90% yield, >20 : 1 dr and >99% ee with chiral squaramide as the catalyst under mild conditions. Notably, this is also the first example of the merger of a spirocyclic pyrazolone scaffold with a spirocyclic thiobutyrolactone scaffold, potentially useful in medicinal chemistry.
Assuntos
Pirazolonas , Ciclização , Pirazolonas/química , Esqueleto , EstereoisomerismoRESUMO
Herein is reported the first example of ring opening and skeletal reconstruction of 3-vinyl benzofuranone-chromones 1 as versatile synthons, which can react with ammonia or primary aliphatic amines as binucleophiles, for the eco-friendly and atom-economical synthesis of diverse and functionalized 2-pyridones 3 with potential biological activity in good to excellent yields (77-93%). When using optically active 1,2-diphenylethylenediamine 2 as the binucleophile, the in situ generated 2-pyridone intermediates are successfully transformed to novel optically active functionalized imidazoline derivatives 4 with high efficiency (up to 87% yield). In particular, this is the first report on the catalyst-free intramolecular cyclization occurring between an amide and a primary aliphatic amine for the construction of imidazoline molecules.
Assuntos
Cromonas , Imidazolinas , Aminas , Catálise , PiridonasRESUMO
BACKGROUND: Assigning every human gene to specific functions, diseases and traits is a grand challenge in modern genetics. Key to addressing this challenge are computational methods, such as supervised learning and label propagation, that can leverage molecular interaction networks to predict gene attributes. In spite of being a popular machine-learning technique across fields, supervised learning has been applied only in a few network-based studies for predicting pathway-, phenotype- or disease-associated genes. It is unknown how supervised learning broadly performs across different networks and diverse gene classification tasks, and how it compares to label propagation, the widely benchmarked canonical approach for this problem. RESULTS: In this study, we present a comprehensive benchmarking of supervised learning for network-based gene classification, evaluating this approach and a classic label propagation technique on hundreds of diverse prediction tasks and multiple networks using stringent evaluation schemes. We demonstrate that supervised learning on a gene's full network connectivity outperforms label propagaton and achieves high prediction accuracy by efficiently capturing local network properties, rivaling label propagation's appeal for naturally using network topology. We further show that supervised learning on the full network is also superior to learning on node embeddings (derived using node2vec), an increasingly popular approach for concisely representing network connectivity. These results show that supervised learning is an accurate approach for prioritizing genes associated with diverse functions, diseases and traits and should be considered a staple of network-based gene classification workflows. AVAILABILITY AND IMPLEMENTATION: The datasets and the code used to reproduce the results and add new gene classification methods have been made freely available. CONTACT: arjun@msu.edu. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.
Assuntos
Biologia Computacional , Redes Reguladoras de Genes , Humanos , Aprendizado de Máquina SupervisionadoRESUMO
Reaching the quantum optics limit of strong light-matter interactions between a single exciton and a plasmon mode is highly desirable, because it opens up possibilities to explore room-temperature quantum devices operating at the single-photon level. However, two challenges severely hinder the realization of this limit: the integration of single-exciton emitters with plasmonic nanostructures and making the coupling strength at the single-exciton level overcome the large damping of the plasmon mode. Here, we demonstrate that these two hindrances can be overcome by attaching individual J aggregates to single cuboid Au@Ag nanorods. In such hybrid nanosystems, both the ultrasmall mode volume of â¼71 nm^{3} and the ultrashort interaction distance of less than 0.9 nm make the coupling coefficient between a single J-aggregate exciton and the cuboid nanorod as high as â¼41.6 meV, enabling strong light-matter interactions to be achieved at the quantum optics limit in single open plasmonic nanocavities.
RESUMO
The polyvinyl alcohol (PVA) protected silver glass-like nanostructure (PVA-Ag-GNS) with high surface-enhanced Raman scattering (SERS) activity was prepared and employed to detect the near-infrared surface enhanced Raman scattering (NIR-SERS) spectra of cytidine aqueous solution (10(-2)-10(-8) mol x L(-1)). In the work, the near-infrared laser beam (785 nm) was used as the excitation light source. The experiment results show that high-quality NIR-SERS spectra were obtained in the ranges of 300 to 2 000 cm(-1) and the detection limit of cytidine aqueous solution was down to 10(-7) mol x L(-1). Meanwhile, the PVA-Ag-GNS shows a high enhancement factor (EF) of -10(8). In order to test the optical reproducibility of PVA-Ag-GNS, ten samples of cytidine aqueous solution (10(-2)-10(-5) mol x L(-1)) had been dropped onto the surface of PVA-Ag-GNS respectively. Meanwhile, these samples were measured by the portable Raman spectrometer. As a result, the PVA-Ag-GNS demonstrated good optical reproducibility in the detection of cytidine aqueous solution. In addition, to explain the reason of enhancement effect, the ultraviolet-visible (UV-Vis) extinction spectrum and scanning electron microscope (SEM) of cytidine molecules adsorbed on the surface of PVA-Ag-GNS were measured. There is plasmon resonance band at 800 nm in the UV-Vis extinction Spectrum of the compound system. Therefore, when the near-infrared laser beam (785 nm) was used as excitation light source, the compound system may produce strongly surface plasmon resonance (SPR). According to the SEM of PVA-Ag-GNS, there are much interstitial between the silver nanoparticles. So NIR-SERS is mainly attributed to electromagnetic (EM) fields associated with strong surface plasmon resonance. At last, the geometry optimization and pre-Raman spectrum of cytidine for the ground states were performed with DFT, B3LYP functional and the 6-311G basis set, and the near-infrared laser with wavelength of 785 nm was employed in the pre-Raman spectrum calculation process. The calculation results without imaginary frequency and the results match pretty well with the experimental Raman spectrum. At the same time, the assignations of Raman bands and adsorption behaviors of cytidine molecules on the surface of PVA-Ag-GNS are also discussed. According to our experiment and calculations, cytidine molecules mainly adsorbed on silver nanoparticles via the ribose moiety and amino group may get close to the local electromagnetic field.
RESUMO
In order to identify volatile organic compounds of fresh plants at room temperature and avoid sample pretreatment and extractions which can be labor intensive, garlic, Chinese chives and scallion were chopped into pieces. Then some of them were placed in the headspace vial and sealed. The gases were drawn from the vial with a syringe and were injected very slowly into Ag colloids for test using R-3000 portable Raman spectrometer. The spectra of volatile organic compounds of allium species, fresh garlic, Chinese chive and shallot plants were successfully.recorded for the first time. For garlic high intensity bands are present at 307, 399, 569, 711, 1,182, 1,287, 1,397 and 1,622 cm(-1). For Chinese chives the high intensity band is present at 672 cm(-1). Low intensity bands are present at 274, 412, 575, 1,185, 1,289, 1,396, 1,618 cm(-1). For shallot high intensity bands are present at 693 cm(-1). Lower intensity bands are present at 372, 888, 1,023 cm(-1). Low intensity bands are present at 1,088, 1,211 and 1,322 cm(-1). The SERS of diallyl disulfide, allyl methyl sulfide and 1-propanethiol in liquid state and gas state were also obtained. The main volatile organic compound of fresh garlic, Chinese chive and shallot are diallyl disulfide, allyl methyl sulfide and 1-propanethiol respectively, and the volatile organic compound of fresh onion, scallion, shallot and chive are all 1-propanethiol. The presented results illustrate that combining headspace and SERS is a powerful tool for volatile organic compound analysis in fresh plants. The volatile organic compound can be detected in fresh plant samples directly and quickly without extraction.
Assuntos
Allium/química , Compostos Orgânicos Voláteis/análise , Allium/classificação , Compostos Alílicos , Dissulfetos , Análise Espectral Raman , Compostos de Sulfidrila , SulfetosRESUMO
The aim is analyzing genetic reLationship and identifying varieties by detecting DNA differences of three kinds of Nicotiana tabacum L. using Fourier infrared spectrum (FTIR). Results show that DNA FTIR of three kinds of Nicotiana tabacum L. is relatively similar. They all have four obvious characteristic peaks. 1 105 cm(-1) beLongs to symmetrical stretching vibration of phosphodiester bond, 1 250 cm(-1) is unsymmetrical stretching vibration of phosphodiester bond, 1 400 cm(-1) is contributed to glucosidic bond, and 1 622 cm(-1) belongs to C4-C5=C6 stretching vibration of cytosine. DNA FTIR data was handled by smoothing, standardizing, second derivative, principal component analysis and Hierarchical cluster analysis. The standard model of Hierarchical cluster combined with principal component of the second derivative was set up. The correct rate of identification is 100%. Yunyan 87 and K326 were clustered into one by using the model. The distance coefficient is 0.003, and DNA similarity is 99.7%, Hongda was clustered into one by itself. The correct rate of cluster is 100%. The study provides a reference for Nicotiana tabacum L. variety identification and genetic breeding.
Assuntos
Nicotiana/genética , Espectroscopia de Infravermelho com Transformada de Fourier , Análise por Conglomerados , DNA de Plantas/análise , Análise de Componente PrincipalRESUMO
Leaves of nine kinds of plants from three subfamily of Rosaceae were used as materials. Genetic relationship was analyzed and species were identified through studying FTIR of nine kinds of plants. Leaves mainly contain large amounts of carbohydrates, proteins, lipids, nucleic acids and other substances. The peaks of carbohydrates are mainly between 1440 and 775 cm(-1). The vibration peaks of the cellulose and lignin are between 1440 and 1337 cm(-1). The peaks between 1000 and 775 cm(-1) are stretching vibration of ribose. The vibration peaks of protein are between 1620 and 1235 cm(-1). The peak at 1620 cm(-1) is sensitive to C=O stretching vibration of protein amide I. The peak at 1523 cm(-1) is assigned to N-H and C-N stretching vibration of protein amide II. Peaks of lipids mainly appeared between 2930 and 1380 cm(-1). The peak at 2922 cm(-1) is CH2 stretching vibration of fat. The peak at 1732 cm(-1) is C=O stretching vibration of fatty acids. The mark peak of the nucleic acid appears in the region between 1250 and 1000 cm(-1). The peak at 1068 cm(-1) is due to the symmetric stretching vibration of PO(2-) group of the phosphodiester-deoxyribose backbone, and the peak at 1246 cm(-1) is associated to the asymmetric stretch vibration of PO(2-) group. The results showed that the cluster model is established by smoothing, standardizing, the second derivative, principal component analysis and Hierarchical cluster analysis. It is accordant with the traditional classification. The result of cluster shows that Prunus armeniaca L. and Prunus seudocerasus Lindl. were clustered into one (Prunoideae). Potentilla fulgens Wall. Rosa chinensis Jacd and Fragaria ananassa Duchesne var. were clustered into the second (Rosoideae). Pyracantha fortuneana Li, Malus pumila Mill. Eriobotrya bengalensis Hook. f. and Malus hallianna Koehne were clustered into the third (Pomoideae). The correct rate of cluster at subfamily is 100%. The correct rate of cluster at genus is 55.56%. The correct rate of identification is 100% when unknown species waiting for determined were laid into the model of Hierarchical cluster to identify. This study provides a new thought and method for genetic relationship analysis of planst.
Assuntos
Folhas de Planta/química , Rosaceae/classificação , Espectroscopia de Infravermelho com Transformada de Fourier , Carboidratos/análise , Celulose/análise , Análise por Conglomerados , Lignina/análise , Lipídeos/análise , Proteínas de Plantas/análiseRESUMO
The classification of Stropharia rugoso-annulata is currently reliant on manual sorting, which may be subject to bias. To improve the sorting efficiency, automated sorting equipment could be used instead. However, sorting naked mushrooms in real time remains a challenging task due to the difficulty of accurately identifying, locating and sorting large quantities of them simultaneously. Models must be deployable on resource-limited devices, making it challenging to achieve both a high accuracy and speed. This paper proposes the APHS-YOLO (YOLOv8n integrated with AKConv, CSPPC and HSFPN modules) model, which is lightweight and efficient, for identifying Stropharia rugoso-annulata of different grades and seasons. This study includes a complete dataset of runners of different grades in spring and autumn. To enhance feature extraction and maintain the recognition accuracy, the new multi-module APHS-YOLO uses HSFPNs (High-Level Screening Feature Pyramid Networks) as a thin-neck structure. It combines an improved lightweight PConv (Partial Convolution)-based convolutional module, CSPPC (Integration of Cross-Stage Partial Networks and Partial Convolution), with the Arbitrary Kernel Convolution (AKConv) module. Additionally, to compensate for the accuracy loss due to lightweighting, APHS-YOLO employs a knowledge refinement technique during training. Compared to the original model, the optimized APHS-YOLO model uses 57.8% less memory and 62.5% fewer computational resources. It has an FPS (frames per second) of over 100 and even achieves 0.1% better accuracy metrics than the original model. These research results provide a valuable reference for the development of automatic sorting equipment for forest farmers.
RESUMO
Objective: This study was conducted to explore the sex differences in the direct and indirect associations among mental health/suicidality, sleep, and screen time. Methods: Using the 2021 Youth Risk Behavior Survey (YRBS) data, 9408 participants were included in the analyses. The associations of endogenous variables (mental health and suicidality), exogenous variable (screen time), mediator (sleep), and covariates (demographic features and risky behaviors) were analyzed using the Structural Equation Model and "medsem" package, as well as logistic regression and bootstrapping methods. To explore the sex differences in the mediation effect, the Likelihood Ratio Test was used for the multiple-group analysis to compare the unconstrained model with the mediation path-constrained model. Results: Screen time had a significant negative association with sleep duration (female: ß = -.09, p < .001; male: ß = -.04, p < .001), positive relations to mental health problems (female: ß = .10, p < .001; male: ß = .12; p < .001), and suicidality (female: OR: 1.07, 95% CI: 1.02-1.13; male: OR: 1.06, 95% CI: 1.01-1.23). Sleep duration is negatively associated with mental health problems (female: ß = -.16, p < .001; male: ß = -.14; p < .001) and negatively associated with the likelihood of suicidality (female: .89, 95% CI: .85-.94; male: OR: .84, 95% CI: .79-.90). Furthermore, female adolescents had higher frequencies in mental health problems, higher odds ratio in suicidality, shorter sleep duration, and bigger mediation effects of sleep, compared to their male counterparts. Conclusion: Mental health and suicidality were affected differently by screen time and sleep between female and male adolescents. Future research may continue to explore sex differences and their underlying reasons.