GraphTGI: an attention-based graph embedding model for predicting TF-target gene interactions.

MOTIVATION: Interaction between transcription factor (TF) and its target genes establishes the knowledge foundation for biological researches in transcriptional regulation, the number of which is, however, still limited by biological techniques. Existing computational methods relevant to the prediction of TF-target interactions are mostly proposed for predicting binding sites, rather than directly predicting the interactions. To this end, we propose here a graph attention-based autoencoder model to predict TF-target gene interactions using the information of the known TF-target gene interaction network combined with two sequential and chemical gene characters, considering that the unobserved interactions between transcription factors and target genes can be predicted by learning the pattern of the known ones. To the best of our knowledge, the proposed model is the first attempt to solve this problem by learning patterns from the known TF-target gene interaction network. RESULTS: In this paper, we formulate the prediction task of TF-target gene interactions as a link prediction problem on a complex knowledge graph and propose a deep learning model called GraphTGI, which is composed of a graph attention-based encoder and a bilinear decoder. We evaluated the prediction performance of the proposed method on a real dataset, and the experimental results show that the proposed model yields outstanding performance with an average AUC value of 0.8864 +/- 0.0057 in the 5-fold cross-validation. It is anticipated that the GraphTGI model can effectively and efficiently predict TF-target gene interactions on a large scale. AVAILABILITY: Python code and the datasets used in our studies are made available at https://github.com/YanghanWu/GraphTGI.

Heterogeneous graph embedding model for predicting interactions between TF and target gene.

MOTIVATION: Identifying the target genes of transcription factors (TFs) is of great significance for biomedical researches. However, using biological experiments to identify TF-target gene interactions is still time consuming, expensive and limited to small scale. Existing computational methods for predicting underlying genes for TF to target is mainly proposed for their binding sites rather than the direct interaction. To bridge this gap, we in this work proposed a deep learning prediction model, named HGETGI, to identify the new TF-target gene interaction. Specifically, the proposed HGETGI model learns the patterns of the known interaction between TF and target gene complemented with their involvement in different human disease mechanisms. It performs prediction based on random walk for meta-path sampling and node embedding in a skip-gram manner. RESULTS: We evaluated the prediction performance of the proposed method on a real dataset and the experimental results show that it can achieve the average area under the curve of 0.8519 ± 0.0731 in fivefold cross validation. Besides, we conducted case studies on the prediction of two important kinds of TF, NFKB1 and TP53. As a result, 33 and 32 in the top-40 ranking lists of NFKB1 and TP53 were successfully confirmed by looking up another public database (hTftarget). It is envisioned that the proposed HGETGI method is feasible and effective for predicting TF-target gene interactions on a large scale. AVAILABILITY AND IMPLEMENTATION: The source code and dataset are available at https://github.com/PGTSING/HGETGI. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Tunable dielectric metasurfaces by structuring the phase-change material.

Metasurfaces have made great progress in the last decade for generating miniature and integrated optical devices. The optical properties of metasurfaces can be tuned dynamically by integrating with phase-change materials. However, the efficiency of tunable metasurfaces remains a bit low, which is a disadvantage for the realistic applications of metasurfaces. Here, we demonstrate the tunable dielectric metasurfaces by structuring the phase-change material Ge2Sb2Te5. The unit cell of metasurface is composed of several Ge2Sb2Te5 nanopillars with different geometric parameters, and the incident light interacts with different nanopillars at diverse phases of Ge2Sb2Te5, leading to various functions. By elaborately arranging the Ge2Sb2Te5 nanopillars, various tunable optical devices have been realized, including tunable beam steering, reconfigurable metalens and switchable wave plate. The refractive direction, focal length and polarization state can be tuned through the phase transition of Ge2Sb2Te5. The phase-change metasurfaces based on Ge2Sb2Te5 nanostructures could be used in cameras, optical microscopy and adaptive optics.

Switchable transmissive and reflective metadevices based on the phase transition of vanadium dioxide.

Metasurfaces have made great progress in the past decade in generating various planar optical devices. However, most metasurfaces exhibit their functions in either reflection mode or transmission mode, with the other mode unutilized. In this work, we demonstrate switchable transmissive and reflective metadevices by combining metasurfaces with vanadium dioxide. The composite metasurface can work as a transmissive metadevice, with one function for vanadium dioxide in the insulating phase, and is changed to a reflective metadevice with another function for vanadium dioxide in the metallic phase. By carefully designing the structures, the metasurface can be switched from a transmissive metalens to a reflective vortex generator, or between a transmissive beam steering and a reflective quarter-wave plate through the phase transition of vanadium dioxide. The switchable transmissive and reflective metadevices have potential applications in imaging, communication, and information processing.

Association between transcription factor 7-like 2 C/T polymorphism and diabetic retinopathy risk: a meta-analysis.

BACKGROUND: Previous studies have suggested a close association between transcription factor 7-like 2 (TCF7L2) polymorphisms and diabetic retinopathy (DR) susceptibility. However, the published results were inconsistent. This meta-analysis was conducted to review and examine the relationship between TCF7L2 rs7903146 C/T polymorphism and DR risk. MATERIALS AND METHODS: Online databases were searched and the related studies were identified in this meta-analysis. Odds ratios (ORs) and corresponding 95% confidence intervals (CIs) were calculated to examine the statistical power. Moreover, heterogeneity test, sensitivity accumulative analysis and publication bias were conducted to measure the statistical effect. RESULT: 6 studies involving 12,982 subjects were included in this meta-analysis to assess the association between rs7903146 C/T polymorphism and DR susceptibility. The synthetic results indicated that the mutation of rs7903146 C/T polymorphism maybe accompany with an increased risk for DR (T vs. C: OR=1.26, 95%CI=1.00-1.60, P=0.05, I2=83.5%; TT vs. CC: OR=1.79 95%CI=1.12-2.86, P=0.02, I2=80.2%; TT vs. CC+CT: OR=1.62, 95%CI=1.38-1.92, P<0.01, I2=32.3%). Moreover, the subgroup analysis also demonstrated an increasing risk for DR with T mutations in Caucasian descendants. CONCLUSION: The current evidences meta-analysis suggested that the TCF7L2 rs7903146 C/T polymorphism might be play an important role in DR susceptibility.

Qualitative study of the learning and studying process of resident physicians in China.

BACKGROUND: Clinical medical education is essential in physician training. This study developed recommendations for medical residency course design on the basis of the perspectives of learners in China and how they interact with their environment. The central research topic was the professional development and learning process of residents, including the obstacles that hinder and factors that promote their learning, their views on existing teaching methods, interaction between teachers and medical teams, and suggestions for designing future residency training programs. METHODS: This study had a qualitative research design. Interviews were conducted between July and October 2019 with 17 specialist residents and 12 assistant general practitioner residents from the department of education of the hospital. The participants were recruited from Qingyuan People's Hospital in Guangdong Province, China. The interview outlines focused on the following four themes: clinical learning experiences and reflections on learning, experience of interaction with patients, experience of working with other medical personnel, and future learning directions. RESULTS: To overcome challenges in clinical learning, the residents mainly learned from their teachers and focused specifically on their own experiences. Regarding teaching methods and designs in clinical medicine, the residents preferred large-group, small-group, and bedside teaching and reported that bedside teaching enables the resolution of clinical problems, initiates self-learning, and improves diagnostic thinking. They disliked teachers with low teaching motivation or who were reluctant to interact with them and favored teachers who had strong teaching skills and respect for their students. CONCLUSIONS: The residents suggested that clinical and active learning must be the main learning method for developing general medical competencies. Residency training must be conducted in an environment that facilitates residents' learning and meaningful learning activities. The interdependent symbiotic relationships in the education ecosystem can serve as a reference for designing residency courses.

The key regulation of miR-124-3p during reprogramming of primary mouse hepatocytes into insulin-producing cells.

Based on the action of small molecule compounds, the efficiency of differentiation of mouse primary hepatocytes into insulin-producing cells (IPCs) was improved by changing the expression of miR-124-2p. Hepatocytes were transfected with microRNA-124-3p (miR-124-3p) mimic or inhibitor, followed by a chemical-defined culture system for maturation of IPCs. Then, detect the expression of insulin-related genes and protein and insulin secretion of each stage during differentiation. The expression of Foxa2, PDX1, NeuroD, insulin1, and insulin2 in IPCs in the miR-124-3p inhibition expression group was significantly upregulated, while the results were opposite in the miR-124-3p overexpression group. The results of cell immunofluorescence and glucose stimulation in vitro of the miR-124-3p inhibition expression group showed that the expression of insulin, PDX1, and C-peptide was increased, and the differentiation efficiency was higher than those of the control group and overexpression group. The primary mouse hepatocytes were successfully reprogrammed into IPCs by small-molecule compounds. We found that miR-124-3p plays a negative regulatory role in the differentiation of hepatocytes into IPCs in vitro. Inhibition of miR-124-3p expression significantly increased the expression of FOXA2 and PDX1, promoted the differentiation of hepatocytes into IPCs, and increased the induction efficiency.

[Network Meta-analysis of effect of traditional Chinese medicine injections in adjuvant treatment of mycoplasma pneumonia in children].

To assess the clinical efficacy of traditional Chinese medicine injection for adjuvant treatment of mycoplasma pneumoniae pneumonia in children by network Meta-analysis method. We retrieved CNKI, WanFang, CBM, Cochrane Library, PubMed from the establishment to September 2018. Two reviewers independently screened out literatures, extracted data and assessed the methodological quality of included studies. The data were analyzed by Stata 13.0 software. Totally 89 RCTs were included, involving 8 kinds of traditional Chinese medical injections and 8 936 patients. According to the results of network Meta-analysis, the order by the total effective rate from high to low was Huangqi Injection>Xiyanping Injection>Tanreqing Injection>Compound Danshen Injection>Reduning Injection>Yanhuning Injection>Qingkailing Injection>Xixinnao Injection; the order by cooling time from high to low was Reduning Injection> Yanhuning Injection>Qingkailing Injection>Tanreqing Injection>Huangqi Injection>Xiyanping Injection>Xiexinnao Injection>Compound Danshen Injection; the order by the cough disappeared time from high to low was Compound Danshen Injection>Qingkailing Injection>Xiyanping Injection>Huangqi Injection>Yanhuning Injection>Reduning Injection>Tanreqing Injection>Xixinnao Injection; the order by the rales disappearing time from high to low was Qingkailing Injection>Yanhuning Injection>Reduning Injection>Huangqi Injection>Tanreqing Injection>Xiyanping Injection>Xixinnao Injection. The results show that traditional Chinese medicine injection has a significant clinical efficacy in the adjuvant treatment of various symptoms of mycoplasma pneumoniae pneumonia in children. Due to the small sample size, more studies are required to verify the strength of evidence.

Pure magnetic-quadrupole scattering and efficient second-harmonic generation from plasmon-dielectric hybrid nano-antennas.

We theoretically demonstrate that pure magnetic quadrupole (MQ) scattering is achieved via the excitation of anapole modes and Fano resonance in noble metal (Au or Ag) and high refractive index dielectric (AlGaAs) hybrid nano-antennas. In Au-AlGaAs hybrid nano-antennas, electric anapole and magnetic anapole modes are observed, leading to the suppressions of electric and magnetic dipoles. Introducing gain material to AlGaAs nanodisk to increase the strength of electric quadrupole (EQ) Fano resonance leads to the suppression of EQ scattering. Then, ideal MQ scattering is achieved at the wavelength of total scattering cross-section dip. The increase of signal-to-noise ratio of MQ results in the great enhancement of near-field inside AlGaAs nanodisk. Additionally, the strong MQ resonance exhibits great capability for boosting second-harmonic generation by proper mode matching. These findings achieved in subwavelength geometries have important implications for functional metamaterials and nonlinear photonic nanodevices.

Induction of hepatocytes-derived insulin-producing cells using small molecules and identification of microRNA profiles during this procedure.

The transplantation of insulin-producing cells (IPCs) or pancreatic progenitor cells is a theoretical therapy for diabetes with insulin insufficiency. Isolated hepatocytes from newborn rats (within 24 h after birth) were progressively induced into IPCs using 5-aza-2'-deoxycytidine, Trichostatin A, retinoic acid, insulin-transferrin-selenium, and nicotinamide. We transplanted Pdx1+ pancreatic progenitors into STZ-induced diabetic mice and found the decreased blood glucose and increased insulin level in comparison with diabetic model. The dynamic expression profiles of microRNAs (miRNAs) were identified using microarray. We found 67 miRNAs were decreasingly expressed; 52 miRNAs were increasingly expressed; 27 miRNAs were specially inhibited in Stage 1 cells (multipotent progenitor cells); and 58 miRNAs were specially inhibited in Pdx1+ cells (Stage 2). Further analysis showed these miRNAs' targets were associated with genetic recombination, stem cell pluripotency maintenance, cellular structure reorganization and insulin secretion. Enrichment analysis using KEGG pathway showed the differentiation of IPCs from hepatocytes was massively more likely not mediated by canonical Wnt/ß-catenin signaling. In addition, the BMP/Smad signaling was involved in this progression. We found the dysregulated miRNAs profiles were inconsistent with cell phenotypes and might be responsible for small molecule-mediated cell differentiation during IPCs induction.

Effects of Passive Smoking on Glycemic Parameters and Lipid Profiles in a Chinese Female Population.

BACKGROUND: The adverse effects of exposure to second-hand cigarette smoke on glucose and lipid parameters in women and the underlying mechanisms are not well understood. In this prospective community-based study, we examined the effects of passive smoking on glycemic parameters and lipid profiles in a Chinese female population. METHODS: Of 3197 healthy women enrolled (30 - 75 years), 2082 self-reported passive exposure to smoke (≥ 15 minutes/day, 3 days/week, > 1 year) and 1115 had no smoke exposure (control group). Data was collected via questionnaire, body measurements, and laboratory assays for glycemic parameters and lipid profiles. RESULTS: (1) Women exposed to second-hand smoke had significantly higher fasting plasma glucose, hemoglobin A1c, total cholesterol, triglycerides, low-density lipoprotein cholesterol, and waist-to-hip ratio compared with the control group (p < 0.05), while 2-h plasma glucose, high-density lipoprotein cholesterol levels, and body mass indexes were similar compared with the control group (p > 0.05). (2) Logistic regression analysis showed that after adjusting for potential confounders, passive smoking women increased the risks of hemoglobin A1c and total cholesterol (p < 0.05). CONCLUSIONS: In this population of Chinese women, exposure to second-hand cigarette smoke was associated with adverse effects on glucose and lipid profiles, suggesting an increased risk of diabetes and cardiovascular disease. These findings support the benefit of stopping smoking in the home and implementing no-smoking regulations in public areas in China and other developing countries to prevent diabetes and other chronic diseases.

[Association between S100B gene polymorphisms and hand, foot and mouth disease caused by enterovirus 71 infection].

OBJECTIVE: To investigate the association between rs9722 polymorphisms in the S100B gene and hand, foot and mouth disease (HFMD) caused by enterovirus 71. METHODS: A total of 124 HFMD children with enterovirus 71 infection were enrolled as subjects, and 56 healthy children were enrolled as control group. The rs9722 polymorphisms in the S100B gene were detected for both groups, and the serum level of S100B protein was measured for 74 HFMD children. RESULTS: The rs9722 locus of the S100B gene had three genotypes, CC, CT, and TT, and the genotype frequencies were in accordance with Hardy-Weinberg equilibrium. Compared with the control group, the HFMD group had significant increases in the frequencies of TT genotype and T allele (P<0.01). Children with severe HFMD caused by enterovirus 71 infection had significantly higher frequencies of TT genotype and T allele than those with moderate or mild HFMD (P<0.05). Compared with the cured patients, the patients with poor prognosis had significant increases in the frequencies of TT genotype and T allele in the rs9722 locus of the S100B gene (P<0.05). Among the 74 children with HFMD, the children with TT genotype had the highest serum level of S100B protein, and those with CC genotype had the lowest level (P<0.01). CONCLUSIONS: T allele in the rs9722 locus of the S100B gene might be a risk factor for severe HFMD caused by enterovirus 71 infection.

[Effect of maternal high-fat diet before and during pregnancy on bone growth of neonatal offspring rats].

OBJECTIVE: To explore the mechanism and effect of maternal high-fat diet before and during pregnancy on bone growth of neonatal offspring rats. METHODS: Forty female Sprague-Dawley rats were divided into high-fat diet and control groups (n=20) that were fed with 35% high-fat diet and standard chow, respectively. After 8 weeks, 8 female rats from each group were sacrificed for liver pathological examinations and the other female rats were mated with male rats and fed continuously with 35% high-fat diet and standard chow throughout gestation, respectively. The body lengths (from apex nasi to end of tail) of the offspring rats from both groups were measured within 24 hours after birth. Enzyme-linked immunosorbent assay was used to detect serum insulin-like growth factor (IFG-I) levels. Liver pathological changes were observed under a light microscope. The expression of insulin receptor substrate 1 (IRS-1) and phosphorylation IRS-1 (Phospho-IRS-1) in tibia and femur samples were detected by immunohistochemistry. The expression of mitogen-activated protein kinase (MAPK) and phosphorylation MAPK (Phospho-MAPK), phosphatidylinositol 3-kinase (PI3K) and phosphorylation PI3K (Phospho-PI3K), protein kinase B (AKT1) and phosphorylation AKT1 (Phospho-AKT1) in tibia and femur samples were detected by Western blot. RESULTS: The offspring rats from the high-fat diet group showed a significant shorter body length compared with those from the control group (P<0.05). The level of serum IGF-I in offspring rats from the high-fat diet group decreased by 20.1% in comparison to those from the control group, but there was no significant difference between the two groups (P>0.05). Fatty degeneration was found in livers of both high-fat diet-fed maternal rats and their offspring rats under a light microscope. There were no significant differences in IRS-1 and Phospho-IRS-1 expression in chondrocytes of tibia and femur samples between the offspring rats of the two groups (P>0.05). The protein expression of MAPK in chondrocytes of tibia and femur samples of offspring rats from the high-fat diet group was higher than that from the control group (P<0.05). There were no significant differences of PI3K and AKT1/Phospho-AKT1 between the offspring rats of the two groups (P>0.05). CONCLUSIONS: A maternal high-fat diet before and during pregnancy may affect the bone growth of offspring rats in utero, which is possibly associated with the decreased IGF-I level. However, further study on the exact mechanism of IGF-I on the bone growth is needed.

A new phenolic amide glycoside from Cimicifuga dahurica.

A new phenolic amide glycoside, cimicifugamide A (1) along with four known compounds, trans-feruloyl tyramine 4-O-beta-D-glucopyranoside (2), (+)-isolariciresinol 3-O-beta-D-glucopyranoside (3), cimidahurine (4), and 24-epi-7, 8-didehydrocimigenol-3-O-beta-D-xylopyranoside (5) were isolated from the rhizomes of Cimicifuga dahurica. Compound 3 was identified as a lignan and has been obtained from Cimicifuga genus for the first time. The structure of compound 1 was elucidated by IR, UV, HR-MS and NMR spectroscopic methods.

Effects of Polyethylene Glycol/Porous Silica Form-Stabilized Phase Change Materials on the Performance of Asphalt Binders.

The road performance and temperature-regulating properties of asphalt binders modified with novel polyethylene glycol (PEG)/porous silica (PS) form-stabilized phase-change materials (PEG/PS-fs-PCMs) were studied. PS and PEG were used as the supporting substance and PCMs. The results showed that PEG/PS-fs-PCMs could maintain a maximum weight percentage of 70% without leakage, at temperatures as high as 90 °C. The PEG/PS-fs-PCMs exhibited stable chemical structures, excellent thermal stability, high heat storage density, and suitable phase-change temperature. Based on conventional physical tests, the addition of PEG/PS-fs-PCMs can increase the viscosity and the degree of hardness of asphalt binders; thus, achieving an excellent comprehensive performance of the modified asphalt binder depends on determining the optimal dosage of PEG/PS-fs-PCMs. Additionally, incorporating PEG/PS-fs-PCM particles into the asphalt binder can enhance its ability to withstand permanent deformation at elevated temperatures, while PEG/PS-fs-PCMs mainly act as a filler, weakening the cohesive force of the asphalt molecules, and preventing the ductility of asphalt from expansion, according to DSR and BBR tests. Moreover, the use of PEG/PS-fs-PCMs can enhance the heat transfer properties of the asphalt binders, resulting in an improved temperature regulation performance. However, the accumulation of PEG/PS-fs-PCM particles on asphalt binders can negatively impact the storage stability of the modified asphalt binders, because of the difference in density between the two materials.

Identification of Network Topology Variations Based on Spectral Entropy.

Based on the fact that the traditional probability distribution entropy describing a local feature of the system cannot effectively capture the global topology variations of the network, some indicators constructed by the network adjacency matrix and Laplacian matrix come into being. Specifically, these measures are based on the eigenvalues of the scaled Laplace matrix, the eigenvalues of the network communicability matrix, and the spectral entropy based on information diffusion that has been proposed recently, respectively. In this article, we systematically study the dependence of these measures on the topological structure of the network. We prove from various aspects that spectral entropy has a better ability to identify the global topology than the traditional distribution entropy. Furthermore, the indicator based on the eigenvalues of the network communicability matrix achieves good results in some aspects while, overall, the spectral entropy is able to identify network topology variations from a global perspective.

Ultra-Broadband, Omnidirectional, High-Efficiency Metamaterial Absorber for Capturing Solar Energy.

In this study, we investigated an absorber based on a center-aligned tandem nanopillar array for ultra-broadband solar energy harvesting theoretically. A high-efficiency, omnidirectional absorber was obtained by introducing the center-aligned tandem nanopillar array embedded in an Al2O3 dielectric layer. The multi-coupling modes at different wavelengths were interpreted. The strong absorption can be adjusted by changing the radii and heights of nanopillars. According to the simulation results, the average absorptance of the absorber exceeded 94% in the wavelength range from 300 nm to 2000 nm. In addition, the high-efficiency absorption was insensitive to the incident angle and polarization state. The research not only proposed an absorber which possesses a huge potential value for application areas, such as thermal photovoltaic systems, infrared detection, and isotropic absorption sensors, but also pointed out a new way to design an absorber with high efficiency in an ultrabroad wavelength range.

[Characteristics of thermal infrared hyperspectra and prediction of sand content of sandy soil].

To explore the potential of thermal infrared hyperspecra for retrieving sand content in soil, the sandy soil was measured using a 102F Fourier Transform Infrared Spectroradiometer (FTIR), and the characteristics of sandy soil's emissivity spectra were discussed based on correlation analysis and principal component analysis. Moreover, the sand contents were predicted using two modeling methods: Partial least squares regression (PLSR) and principal component regression (PCR). The results show that the Reststrahlen feature (RF) of SiO2 is obvious in the emissivity spectra of sandy soil with two large asymmetrical absorption troughs near 8.13 and 9.17 microm and two small troughs in the region of 12-13 microm. Soil emissivity becomes lower when sand content increases, this trend is more evident especially in the regions of 8-9.5 microm and 9.5-10.4 microm of which correlation coefficients are above 0.65 and 0.5 respectively, and these two regions can account for 84.07% of total emissivity variance. Predictive precision varies significantly when sand content is predicted by different modeling methods or spectral variables. The PLSR model can achieve the highest predictive precision by using first-order derivative spectra, and it's RMSE of modeling and prediction is 0.45 and 0.53 respectively, and the R2, 0.9907 and 0.9836, which means that the thermal hyperspectra has promising potential for retrieving sand content in soil.

Finite-size scaling of geometric renormalization flows in complex networks.

Some characteristics of complex networks need to be derived from global knowledge of the network topologies, which challenges the practice for studying many large-scale real-world networks. Recently, the geometric renormalization technique has provided a good approximation framework to significantly reduce the size and complexity of a network while retaining its "slow" degrees of freedom. However, due to the finite-size effect of real networks, excessive renormalization iterations will eventually cause these important "slow" degrees of freedom to be filtered out. In this paper, we systematically investigate the finite-size scaling of structural and dynamical observables in geometric renormalization flows of both synthetic and real evolutionary networks. Our results show that these observables can be well characterized by a certain scaling function. Specifically, we show that the critical exponent implied by the scaling function is independent of these observables but depends only on the structural properties of the network. To a certain extent, the results of this paper are of great significance for predicting the observable quantities of large-scale real systems and further suggest that the potential scale invariance of many real-world networks is often masked by finite-size effects.

Non-Causal Effects of Asthma on COVID-19 Susceptibility and Severity.

Background: Asthma is observationally associated with an increased risk of COVID-19, but the causality remains unclear. We aim to determine whether there is a casual role of asthma in susceptibility to SARS-CoV-2 infection or COVID-19 severity. Methods: Instrumental variables (IVs) for asthma and moderate-to-severe asthma were obtained from publicly available summary statistics from the most recent and largest genome-wide association study (GWAS), including 394 283 and 57 695 participants of European ancestry, respectively. The corresponding data for COVID-19 susceptibility, hospitalization and severe-disease were derived from the COVID-19 Host Genetics Initiative GWAS meta-analysis of up to 1 683 768 individuals of European descent. Causality was inferred between correlated traits by Mendelian Randomization analyses. Inverse-variance weighted method was used as the primary MR estimates and multiple alternate approaches and several sensitivity analyses were also conducted. Results: Our MR analysis revealed no causal effects of asthma on COVID-19 susceptibility, hospitalization or severe disease, with odds ratio (OR) of 0.994 (95% CI: 0.962-1.027), 1.020 (95% CI: 0.955-1.089), and 0.929 (95% CI: 0.836-1.032), respectively. Furthermore, using genetic variants for moderate-to-severe asthma, a similar pattern of results was observed for COVID-19 susceptibility (OR: 0.988, 95% CI: 0.946-1.031), hospitalization (OR: 0.967, 95% CI: 0.906-1.031), and severe disease (OR: 0.911, 95% CI: 0.823-1.009). The association of asthma and moderate-to-severe asthma with COVID-19 was overall robust to sensitivity analyses. Conclusion: Genetically predicted asthma was not associated with susceptibility to, or severity of, COVID-19 disease, indicating that asthma is unlikely to be a causal factor in the development of COVID-19.