Association of COVID-19 vaccination with risks of hospitalization due to cardiovascular and other diseases: A study using data from the UK Biobank.

OBJECTIVES: To explore whether COVID-19 vaccination protects against hospital admission by preventing infections and severe disease. METHODS: We leveraged the UK Biobank and studied associations of COVID-19 vaccination (BioNTech-BNT162b2 or Oxford-AstraZeneca-ChAdOx1) with hospitalizations from cardiovascular and other selected diseases (N = 393,544; median follow-up = 54 days among vaccinated individuals). Multivariable Cox, Poisson regression, propensity score matching, and inverse probability treatment weighting analyses were performed. We also performed adjustment using prescription-time distribution matching, and prior event rate ratio. RESULTS: We observed that COVID-19 vaccination (at least one dose), compared with no vaccination, was associated with reduced short-term risks of hospitalizations from stroke (hazard ratio [HR] = 0.178, 95% confidence interval [CI]: 0.127-0.250, P = 1.50e-23), venous thromboembolism (HR = 0.426, CI: 0.270-0.673, P = 2.51e-4), dementia (HR = 0.114, CI: 0.060-0.216; P = 2.24e-11), non-COVID-19 pneumonia (HR = 0.108, CI: 0.080-0.145; P = 2.20e-49), coronary artery disease (HR = 0.563, CI: 0.416-0.762; P = 2.05e-4), chronic obstructive pulmonary disease (HR = 0.212, CI: 0.126-0.357; P = 4.92e-9), type 2 diabetes (HR = 0.216, CI: 0.096-0.486, P = 2.12e-4), heart failure (HR = 0.174, CI: 0.118-0.256, P = 1.34e-18), and renal failure (HR = 0.415, CI: 0.255-0.677, P = 4.19e-4), based on standard Cox regression models. Among the previously mentioned results, reduced hospitalizations for stroke, heart failure, non-COVID-19 pneumonia, and dementia were consistently observed across regression, propensity score matching/inverse probability treatment weighting, prescription-time distribution matching, and prior event rate ratio. The results for two-dose vaccination were similar. CONCLUSIONS: Taken together, this study provides further support to the safety and benefits of COVID-19 vaccination, and such benefits may extend beyond reduction of infection risk or severity per se. However, causal relationship cannot be concluded and further studies are required.

Whole-exome sequencing in a Chinese sample provides preliminary evidence for the link between rare/low-frequency immune-related variants and early-onset schizophrenia.

Rare and low-frequency variants contribute to schizophrenia (SCZ), and may influence its age-at-onset (AAO). We examined the association of rare or low-frequency deleterious coding variants in Chinese patients with SCZ. We collected DNA samples in 197 patients with SCZ spectrum disorder and 82 healthy controls (HC), and performed exome sequencing. The AAO variable was ascertained in the majority of SCZ participants for identify the early-onset (EOS, AAO<=18) and adult-onset (AOS, AAO>18) subgroups. We examined the overall association of rare/low-frequency, damaging variants in SCZ versus HC, EOS versus HC, and AOS versus HC at the gene and gene-set levels using Sequence Kernel Association Test. The quantitative rare-variant association test of AAO was conducted. Resampling was used to obtain empirical p-values and to control for family-wise error rate (FWER). In binary-trait association tests, we identified 5 potential candidate risk genes and 10 gene ontology biological processes (GOBP) terms, among which PADI2 reached FWER-adjusted significance. In quantitative rare-variant association tests, we found marginally significant correlations of AAO with alterations in 4 candidate risk genes, and 5 GOBP pathways. Together, the biological and functional profiles of these genes and gene sets supported the involvement of perturbations of neural systems in SCZ, and altered immune functions in EOS.

A genome-wide association study of Chinese and English language phenotypes in Hong Kong Chinese children.

Dyslexia and developmental language disorders are important learning difficulties. However, their genetic basis remains poorly understood, and most genetic studies were performed on Europeans. There is a lack of genome-wide association studies (GWAS) on literacy phenotypes of Chinese as a native language and English as a second language (ESL) in a Chinese population. In this study, we conducted GWAS on 34 reading/language-related phenotypes in Hong Kong Chinese bilingual children (including both twins and singletons; total N = 1046). We performed association tests at the single-variant, gene, and pathway levels. In addition, we tested genetic overlap of these phenotypes with other neuropsychiatric disorders, as well as cognitive performance (CP) and educational attainment (EA) using polygenic risk score (PRS) analysis. Totally 5 independent loci (LD-clumped at r2 = 0.01; MAF > 0.05) reached genome-wide significance (p < 5e-08; filtered by imputation quality metric Rsq>0.3 and having at least 2 correlated SNPs (r2 > 0.5) with p < 1e-3). The loci were associated with a range of language/literacy traits such as Chinese vocabulary, character and word reading, and rapid digit naming, as well as English lexical decision. Several SNPs from these loci mapped to genes that were reported to be associated with EA and other neuropsychiatric phenotypes, such as MANEA and PLXNC1. In PRS analysis, EA and CP showed the most consistent and significant polygenic overlap with a variety of language traits, especially English literacy skills. To summarize, this study revealed the genetic basis of Chinese and English abilities in a group of Chinese bilingual children. Further studies are warranted to replicate the findings.

INSIDER: Interpretable sparse matrix decomposition for RNA expression data analysis.

RNA sequencing (RNA-Seq) is widely used to capture transcriptome dynamics across tissues, biological entities, and conditions. Currently, few or no methods can handle multiple biological variables (e.g., tissues/ phenotypes) and their interactions simultaneously, while also achieving dimension reduction (DR). We propose INSIDER, a general and flexible statistical framework based on matrix factorization, which is freely available at https://github.com/kai0511/insider. INSIDER decomposes variation from different biological variables and their interactions into a shared low-rank latent space. Particularly, it introduces the elastic net penalty to induce sparsity while considering the grouping effects of genes. It can achieve DR of high-dimensional data (of > = 3 dimensions), as opposed to conventional methods (e.g., PCA/NMF) which generally only handle 2D data (e.g., sample × expression). Besides, it enables computing 'adjusted' expression profiles for specific biological variables while controlling variation from other variables. INSIDER is computationally efficient and accommodates missing data. INSIDER also performed similarly or outperformed a close competing method, SDA, as shown in simulations and can handle complex missing data in RNA-Seq data. Moreover, unlike SDA, it can be used when the data cannot be structured into a tensor. Lastly, we demonstrate its usefulness via real data analysis, including clustering donors for disease subtyping, revealing neuro-development trajectory using the BrainSpan data, and uncovering biological processes contributing to variables of interest (e.g., disease status and tissue) and their interactions.

Bidirectional two-sample Mendelian randomization study of differential white blood cell counts and schizophrenia.

BACKGROUND: Schizophrenia and white blood cell counts (WBC) are both complex and polygenic traits. Previous evidence suggests that increased WBC are associated with higher all-cause mortality, and other studies have found elevated WBC in first-episode psychosis and chronic schizophrenia. However, these observational findings may be confounded by antipsychotic exposures and their effects on WBC. Mendelian randomization (MR) is a useful method for examining the directions of genetically-predicted relationships between schizophrenia and WBC. METHODS: We performed a two-sample MR using summary statistics from genome-wide association studies (GWAS) conducted by the Psychiatric Genomics Consortium Schizophrenia Workgroup (N = 130,644) and the Blood Cell Consortium (N = 563,946). The MR methods included inverse variance weighted (IVW), MR Egger, weighted median, MR-PRESSO, contamination mixture, and a novel approach called mixture model reciprocal causal inference (MRCI). False discovery rate was employed to correct for multiple testing. RESULTS: Multiple MR methods supported bidirectional genetically-predicted relationships between lymphocyte count and schizophrenia: IVW (b = 0.026; FDR p-value = 0.008), MR Egger (b = 0.026; FDR p-value = 0.008), weighted median (b = 0.013; FDR p-value = 0.049), and MR-PRESSO (b = 0.014; FDR p-value = 0.010) in the forward direction, and IVW (OR = 1.100; FDR p-value = 0.021), MR Egger (OR = 1.231; FDR p-value < 0.001), weighted median (OR = 1.136; FDR p-value = 0.006) and MRCI (OR = 1.260; FDR p-value = 0.026) in the reverse direction. MR Egger (OR = 1.171; FDR p-value < 0.001) and MRCI (OR = 1.154; FDR p-value = 0.026) both suggested genetically-predicted eosinophil count is associated with schizophrenia, but MR Egger (b = 0.060; FDR p-value = 0.010) and contamination mixture (b = -0.013; FDR p-value = 0.045) gave ambiguous results on whether genetically predicted liability to schizophrenia would be associated with eosinophil count. MR Egger (b = 0.044; FDR p-value = 0.010) and MR-PRESSO (b = 0.009; FDR p-value = 0.045) supported genetically predicted liability to schizophrenia is associated with elevated monocyte count, and the opposite direction was also indicated by MR Egger (OR = 1.231; FDR p-value = 0.045). Lastly, unidirectional genetic liability from schizophrenia to neutrophil count were proposed by MR-PRESSO (b = 0.011; FDR p-value = 0.028) and contamination mixture (b = 0.011; FDR p-value = 0.045) method. CONCLUSION: This MR study utilised multiple MR methods to obtain results suggesting bidirectional genetic genetically-predicted relationships for elevated lymphocyte counts and schizophrenia risk. In addition, moderate evidence also showed bidirectional genetically-predicted relationships between schizophrenia and monocyte counts, and unidirectional effect from genetic liability for eosinophil count to schizophrenia and from genetic liability for schizophrenia to neutrophil count. The influence of schizophrenia to eosinophil count is less certain. Our findings support the role of WBC in schizophrenia and concur with the hypothesis of neuroinflammation in schizophrenia.

SumVg: Total Heritability Explained by All Variants in Genome-Wide Association Studies Based on Summary Statistics with Standard Error Estimates.

Genome-wide association studies (GWAS) are commonly employed to study the genetic basis of complex traits/diseases, and a key question is how much heritability could be explained by all single nucleotide polymorphisms (SNPs) in GWAS. One widely used approach that relies on summary statistics only is linkage disequilibrium score regression (LDSC); however, this approach requires certain assumptions about the effects of SNPs (e.g., all SNPs contribute to heritability and each SNP contributes equal variance). More flexible modeling methods may be useful. We previously developed an approach recovering the "true" effect sizes from a set of observed z-statistics with an empirical Bayes approach, using only summary statistics. However, methods for standard error (SE) estimation are not available yet, limiting the interpretation of our results and the applicability of the approach. In this study, we developed several resampling-based approaches to estimate the SE of SNP-based heritability, including two jackknife and three parametric bootstrap methods. The resampling procedures are performed at the SNP level as it is most common to estimate heritability from GWAS summary statistics alone. Simulations showed that the delete-d-jackknife and parametric bootstrap approaches provide good estimates of the SE. In particular, the parametric bootstrap approaches yield the lowest root-mean-squared-error (RMSE) of the true SE. We also explored various methods for constructing confidence intervals (CIs). In addition, we applied our method to estimate the SNP-based heritability of 12 immune-related traits (levels of cytokines and growth factors) to shed light on their genetic architecture. We also implemented the methods to compute the sum of heritability explained and the corresponding SE in an R package SumVg. In conclusion, SumVg may provide a useful alternative tool for calculating SNP heritability and estimating SE/CI, which does not rely on distributional assumptions of SNP effects.

Early-onset schizophrenia is associated with immune-related rare variants in a Chinese sample.

Background: Rare variants are likely to contribute to schizophrenia (SCZ), given the large discrepancy between the heritability estimated from twin and GWAS studies. Furthermore, the nature of the rare-variant contribution to SCZ may vary with the "age-at-onset" (AAO), since early-onset has been suggested as being indicative of neurodevelopment deviance. Objective: To examine the association of rare deleterious coding variants in early- and adult-onset SCZ in a Chinese sample. Method: Exome sequencing was performed on DNA from 197 patients with SCZ spectrum disorder and 82 healthy controls (HC) of Chinese ancestry recruited in Hong Kong. We also gathered AAO information in the majority of SCZ samples. Patients were classified into early-onset (EOS, AAO<18) and adult-onset (AOS, AAO>18). We collapsed the rare variants to improve statistical power and examined the overall association of rare variants in SCZ versus HC, EOS versus HC, and AOS versus HC at the gene and gene-set levels by Sequence Kernel Association Test. The quantitative rare-variant association test of AAO was also conducted. We focused on variants which were predicted to have a medium or high impact on the protein-encoding process as defined by Ensembl. We applied a 100000-time permutation test to obtain empirical p-values, with significance threshold set at p < 1e -3 to control family-wise error rates. Moreover, we compared the burden of targeted rare variants in significant risk genes and gene sets in cases and controls. Results: Based on several binary-trait association tests (i.e., SCZ vs HC, EOS vs HC and AOS vs HC), we identified 7 candidate risk genes and 20 gene ontology biological processes (GOBP) terms, which exhibited higher burdens in SCZ than in controls. Based on quantitative rare-variant association tests, we found that alterations in 5 candidate risk genes and 7 GOBP pathways were significantly correlated with AAO. Based on biological and functional profiles of the candidate risk genes and gene sets, our findings suggested that, in addition to the involvement of perturbations in neural systems in SCZ in general, altered immune responses may be specifically implicated in EOS. Conclusion: Disrupted immune responses may exacerbate abnormal perturbations during neurodevelopment and trigger the early onset of SCZ. We provided evidence of rare variants increasing SCZ risk in the Chinese population.

Pathway-Specific Polygenic Scores Improve Cross-Ancestry Prediction of Psychosis and Clinical Outcomes.

Psychotic disorders are debilitating conditions with disproportionately high public health burden. Genetic studies indicate high heritability, but current polygenic scores (PGS) account for only a fraction of variance in psychosis risk. PGS often show poor portability across ancestries, performing significantly worse in non-European populations. Pathway-specific PGS (pPGS), which restrict PGS to genomic locations within distinct biological units, could lead to increased mechanistic understanding of pathways that lead to risk and improve cross-ancestry prediction by reducing noise in genetic predictors. This study examined the predictive power of genome-wide PGS and nine pathway-specific pPGS in a unique Chinese-ancestry sample of deeply-phenotyped psychosis patients and non-psychiatric controls. We found strong evidence for the involvement of schizophrenia-associated risk variants within "nervous system development" (p=2.5e-4) and "regulation of neuron differentiation" pathways (p=3.0e-4) in predicting risk for psychosis. We also found the "ion channel complex" pPGS, with weights derived from GWAS of bipolar disorder, to be strongly associated with the number of inpatient psychiatry admissions a patient experiences (p=1.5e-3) and account for a majority of the signal in the overall bipolar PGS. Importantly, although the schizophrenia genome-wide PGS alone explained only 3.7% of the variance in liability to psychosis in this Chinese ancestry sample, the addition of the schizophrenia-weighted pPGS for "nervous system development" and "regulation of neuron differentiation" increased the variance explained to 6.9%, which is on-par with the predictive power of PGS in European ancestry samples. Thus, not only can pPGS provide greater insight into mechanisms underlying genetic risk for disease and clinical outcomes, but may also improve cross-ancestry risk prediction accuracy.

Dyslexia-related loci are significantly associated with language and literacy in Chinese-English bilingual Hong Kong Chinese twins.

A recent genome-wide association study on dyslexia in 51,800 affected European adults and 1,087,070 controls detected 42 genome-wide significant single nucleotide variants (SNPs). The association between rs2624839 in SEMA3F and reading fluency was replicated in a Chinese cohort. This study explores the genetic overlap between Chinese and English word reading, vocabulary knowledge and spelling, and aims at replicating the association in a unique cohort of bilingual (Chinese-English) Hong Kong Chinese twins. Our result showed an almost complete genetic overlap in vocabulary knowledge (r2 = 0.995), and some genetic overlaps in word reading and spelling (r2 = 0.846, 0.687) across the languages. To investigate the region near rs2624839, we tested proxy SNPs (rs1005678, rs12632110 and rs12494414) at the population level (n = 305-308) and the within-twin level (n = 342-344 [171-172 twin pairs]). All the three SNPs showed significant associations with quantitative Chinese and English vocabulary knowledge (p < 0.05). The strongest association after multiple testing correction was between rs12494414 and English vocabulary knowledge at the within-twin level (p = 0.004). There was a trend of associations with word reading and spelling in English but not in Chinese. Our result suggested that the region near rs2624839 is one of the common genetic factors across English and Chinese vocabulary knowledge and unique factors of English word reading and English spelling in bilingual Chinese twins. A larger sample size is required to validate our findings. Further studies on the relationship between variable expression of SEMA3F, which is important to neurodevelopment, and language and literacy are encouraged.

The genetic basis of onset age in schizophrenia: evidence and models.

Schizophrenia is a heritable neurocognitive disorder affecting about 1% of the population, and usually has an onset age at around 21-25 in males and 25-30 in females. Recent advances in genetics have helped to identify many common and rare variants for the liability to schizophrenia. Earlier evidence appeared to suggest that younger onset age is associated with higher genetic liability to schizophrenia. Clinical longitudinal research also found that early and very-early onset schizophrenia are associated with poor clinical, neurocognitive, and functional profiles. A recent study reported a heritability of 0.33 for schizophrenia onset age, but the genetic basis of this trait in schizophrenia remains elusive. In the pre-Genome-Wide Association Study (GWAS) era, genetic loci found to be associated with onset age were seldom replicated. In the post-Genome-Wide Association Study era, new conceptual frameworks are needed to clarify the role of onset age in genetic research in schizophrenia, and to identify its genetic basis. In this review, we first discussed the potential of onset age as a characterizing/subtyping feature for psychosis, and as an important phenotypic dimension of schizophrenia. Second, we reviewed the methods, samples, findings and limitations of previous genetic research on onset age in schizophrenia. Third, we discussed a potential conceptual framework for studying the genetic basis of onset age, as well as the concepts of susceptibility, modifier, and "mixed" genes. Fourth, we discussed the limitations of this review. Lastly, we discussed the potential clinical implications for genetic research of onset age of schizophrenia, and how future research can unveil the potential mechanisms for this trait.

The single-cell chromatin accessibility landscape in mouse perinatal testis development.

Spermatogenesis depends on an orchestrated series of developing events in germ cells and full maturation of the somatic microenvironment. To date, the majority of efforts to study cellular heterogeneity in testis has been focused on single-cell gene expression rather than the chromatin landscape shaping gene expression. To advance our understanding of the regulatory programs underlying testicular cell types, we analyzed single-cell chromatin accessibility profiles in more than 25,000 cells from mouse developing testis. We showed that single-cell sequencing assay for transposase-accessible chromatin (scATAC-Seq) allowed us to deconvolve distinct cell populations and identify cis-regulatory elements (CREs) underlying cell-type specification. We identified sets of transcription factors associated with cell type-specific accessibility, revealing novel regulators of cell fate specification and maintenance. Pseudotime reconstruction revealed detailed regulatory dynamics coordinating the sequential developmental progressions of germ cells and somatic cells. This high-resolution dataset also unveiled previously unreported subpopulations within both the Sertoli and Leydig cell groups. Further, we defined candidate target cell types and genes of several genome-wide association study (GWAS) signals, including those associated with testosterone levels and coronary artery disease. Collectively, our data provide a blueprint of the 'regulon' of the mouse male germline and supporting somatic cells.

Prediction of Drug Targets for Specific Diseases Leveraging Gene Perturbation Data: A Machine Learning Approach.

Identification of the correct targets is a key element for successful drug development. However, there are limited approaches for predicting drug targets for specific diseases using omics data, and few have leveraged expression profiles from gene perturbations. We present a novel computational approach for drug target discovery based on machine learning (ML) models. ML models are first trained on drug-induced expression profiles with outcomes defined as whether the drug treats the studied disease. The goal is to "learn" the expression patterns associated with treatment. Then, the fitted ML models were applied to expression profiles from gene perturbations (overexpression (OE)/knockdown (KD)). We prioritized targets based on predicted probabilities from the ML model, which reflects treatment potential. The methodology was applied to predict targets for hypertension, diabetes mellitus (DM), rheumatoid arthritis (RA), and schizophrenia (SCZ). We validated our approach by evaluating whether the identified targets may 're-discover' known drug targets from an external database (OpenTargets). Indeed, we found evidence of significant enrichment across all diseases under study. A further literature search revealed that many candidates were supported by previous studies. For example, we predicted PSMB8 inhibition to be associated with the treatment of RA, which was supported by a study showing that PSMB8 inhibitors (PR-957) ameliorated experimental RA in mice. In conclusion, we propose a new ML approach to integrate the expression profiles from drugs and gene perturbations and validated the framework. Our approach is flexible and may provide an independent source of information when prioritizing drug targets.

Contributions of common genetic variants to specific languages and to when a language is learned.

Research over the past two decades has identified a group of common genetic variants explaining a portion of variance in native language ability. The present study investigates whether the same group of genetic variants are associated with different languages and languages learned at different times in life. We recruited 940 young adults who spoke from childhood Chinese and English as their first (native) (L1) and second (L2) language, respectively, who were learners of a new, third (L3) language. For the variants examined, we found a general decrease of contribution of genes to language functions from native to foreign (L2 and L3) languages, with variance in foreign languages explained largely by non-genetic factors such as musical training and motivation. Furthermore, genetic variants that were found to contribute to traits specific to Chinese and English respectively exerted the strongest effects on L1 and L2. These results seem to speak against the hypothesis of a language- and time-universal genetic core of linguistic functions. Instead, they provide preliminary evidence that genetic contribution to language may depend at least partly on the intricate language-specific features. Future research including a larger sample size, more languages and more genetic variants is required to further explore these hypotheses.

Shortened Leukocyte Telomere Length Is Associated With Glycemic Progression in Type 2 Diabetes: A Prospective and Mendelian Randomization Analysis.

OBJECTIVE: Several studies support associations between relative leukocyte telomere length (rLTL), a biomarker of biological aging and type 2 diabetes. This study investigates the relationship between rLTL and the risk of glycemic progression in patients with type 2 diabetes. RESEARCH DESIGN AND METHODS: In this cohort study, consecutive Chinese patients with type 2 diabetes (N = 5,506) from the Hong Kong Diabetes Register with stored baseline DNA and available follow-up data were studied. rLTL was measured using quantitative PCR. Glycemic progression was defined as the new need for exogenous insulin. RESULTS: The mean (SD) age of the 5,349 subjects was 57.0 (13.3) years, and mean (SD) follow-up was 8.8 (5.4) years. Baseline rLTL was significantly shorter in the 1,803 subjects who progressed to insulin requirement compared with the remaining subjects (4.43 ± 1.16 vs. 4.69 ± 1.20). Shorter rLTL was associated with a higher risk of glycemic progression (hazard ratio [95% CI] for each unit decrease [to ∼0.2 kilobases]: 1.10 [1.06-1.14]), which remained significant after adjusting for confounders. Baseline rLTL was independently associated with glycemic exposure during follow-up (ß = -0.05 [-0.06 to -0.04]). Each 1-kilobase decrease in absolute LTL was on average associated with a 1.69-fold higher risk of diabetes progression (95% CI 1.35-2.11). Two-sample Mendelian randomization analysis showed per 1-unit genetically decreased rLTL was associated with a 1.38-fold higher risk of diabetes progression (95% CI 1.12-1.70). CONCLUSIONS: Shorter rLTL was significantly associated with an increased risk of glycemic progression in individuals with type 2 diabetes, independent of established risk factors. Telomere length may be a useful biomarker for glycemic progression in people with type 2 diabetes.

Exploring Drugs and Vaccines Associated with Altered Risks and Severity of COVID-19: A UK Biobank Cohort Study of All ATC Level-4 Drug Categories Reveals Repositioning Opportunities.

Effective therapies for COVID-19 are still lacking, and drug repositioning is a promising approach to address this problem. Here, we adopted a medical informatics approach to repositioning. We leveraged a large prospective cohort, the UK-Biobank (UKBB, N ~ 397,000), and studied associations of prior use of all level-4 ATC drug categories (N = 819, including vaccines) with COVID-19 diagnosis and severity. Effects of drugs on the risk of infection, disease severity, and mortality were investigated separately. Logistic regression was conducted, controlling for main confounders. We observed strong and highly consistent protective associations with statins. Many top-listed protective drugs were also cardiovascular medications, such as angiotensin-converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB), calcium channel blocker (CCB), and beta-blockers. Some other drugs showing protective associations included biguanides (metformin), estrogens, thyroid hormones, proton pump inhibitors, and testosterone-5-alpha reductase inhibitors, among others. We also observed protective associations by influenza, pneumococcal, and several other vaccines. Subgroup and interaction analyses were also conducted, which revealed differences in protective effects in various subgroups. For example, protective effects of flu/pneumococcal vaccines were weaker in obese individuals, while protection by statins was stronger in cardiovascular patients. To conclude, our analysis revealed many drug repositioning candidates, for example several cardiovascular medications. Further studies are required for validation.

Uncovering Clinical Risk Factors and Predicting Severe COVID-19 Cases Using UK Biobank Data: Machine Learning Approach.

BACKGROUND: COVID-19 is a major public health concern. Given the extent of the pandemic, it is urgent to identify risk factors associated with disease severity. More accurate prediction of those at risk of developing severe infections is of high clinical importance. OBJECTIVE: Based on the UK Biobank (UKBB), we aimed to build machine learning models to predict the risk of developing severe or fatal infections, and uncover major risk factors involved. METHODS: We first restricted the analysis to infected individuals (n=7846), then performed analysis at a population level, considering those with no known infection as controls (ncontrols=465,728). Hospitalization was used as a proxy for severity. A total of 97 clinical variables (collected prior to the COVID-19 outbreak) covering demographic variables, comorbidities, blood measurements (eg, hematological/liver/renal function/metabolic parameters), anthropometric measures, and other risk factors (eg, smoking/drinking) were included as predictors. We also constructed a simplified (lite) prediction model using 27 covariates that can be more easily obtained (demographic and comorbidity data). XGboost (gradient-boosted trees) was used for prediction and predictive performance was assessed by cross-validation. Variable importance was quantified by Shapley values (ShapVal), permutation importance (PermImp), and accuracy gain. Shapley dependency and interaction plots were used to evaluate the pattern of relationships between risk factors and outcomes. RESULTS: A total of 2386 severe and 477 fatal cases were identified. For analyses within infected individuals (n=7846), our prediction model achieved area under the receiving-operating characteristic curve (AUC-ROC) of 0.723 (95% CI 0.711-0.736) and 0.814 (95% CI 0.791-0.838) for severe and fatal infections, respectively. The top 5 contributing factors (sorted by ShapVal) for severity were age, number of drugs taken (cnt_tx), cystatin C (reflecting renal function), waist-to-hip ratio (WHR), and Townsend deprivation index (TDI). For mortality, the top features were age, testosterone, cnt_tx, waist circumference (WC), and red cell distribution width. For analyses involving the whole UKBB population, AUCs for severity and fatality were 0.696 (95% CI 0.684-0.708) and 0.825 (95% CI 0.802-0.848), respectively. The same top 5 risk factors were identified for both outcomes, namely, age, cnt_tx, WC, WHR, and TDI. Apart from the above, age, cystatin C, TDI, and cnt_tx were among the top 10 across all 4 analyses. Other diseases top ranked by ShapVal or PermImp were type 2 diabetes mellitus (T2DM), coronary artery disease, atrial fibrillation, and dementia, among others. For the "lite" models, predictive performances were broadly similar, with estimated AUCs of 0.716, 0.818, 0.696, and 0.830, respectively. The top ranked variables were similar to above, including age, cnt_tx, WC, sex (male), and T2DM. CONCLUSIONS: We identified numerous baseline clinical risk factors for severe/fatal infection by XGboost. For example, age, central obesity, impaired renal function, multiple comorbidities, and cardiometabolic abnormalities may predispose to poorer outcomes. The prediction models may be useful at a population level to identify those susceptible to developing severe/fatal infections, facilitating targeted prevention strategies. A risk-prediction tool is also available online. Further replications in independent cohorts are required to verify our findings.

Analysis of genetic differences between psychiatric disorders: exploring pathways and cell types/tissues involved and ability to differentiate the disorders by polygenic scores.

Although displaying genetic correlations, psychiatric disorders are clinically defined as categorical entities as they each have distinguishing clinical features and may involve different treatments. Identifying differential genetic variations between these disorders may reveal how the disorders differ biologically and help to guide more personalized treatment. Here we presented a statistical framework and comprehensive analysis to identify genetic markers differentially associated with various psychiatric disorders/traits based on GWAS summary statistics, covering 18 psychiatric traits/disorders and 26 comparisons. We also conducted comprehensive analysis to unravel the genes, pathways and SNP functional categories involved, and the cell types and tissues implicated. We also assessed how well one could distinguish between psychiatric disorders by polygenic risk scores (PRS). SNP-based heritabilities (h2snp) were significantly larger than zero for most comparisons. Based on current GWAS data, PRS have mostly modest power to distinguish between psychiatric disorders. For example, we estimated that AUC for distinguishing schizophrenia from major depressive disorder (MDD), bipolar disorder (BPD) from MDD and schizophrenia from BPD were 0.694, 0.602 and 0.618, respectively, while the maximum AUC (based on h2snp) were 0.763, 0.749 and 0.726, respectively. We also uncovered differences in each pair of studied traits in terms of their differences in genetic correlation with comorbid traits. For example, clinically defined MDD appeared to more strongly genetically correlated with other psychiatric disorders and heart disease, when compared to non-clinically defined depression in UK Biobank. Our findings highlight genetic differences between psychiatric disorders and the mechanisms involved. PRS may help differential diagnosis of selected psychiatric disorders in the future with larger GWAS samples.

Association of Sleep Traits and Heel Bone Mineral Density: Observational and Mendelian Randomization Studies.

Observational studies have suggested that sleep and circadian disturbances are potentially modifiable risk factors for low bone mineral density (BMD), but the causal relationship is unclear. This study aimed to (i) replicate the findings by examining observational association of sleep traits with low estimated BMD); (ii) examine whether these associations were causal by using Mendelian randomization (MR) analyses; and (iii) investigate potential modulation effects of sex and menopause. A total of 398,137 White British subjects (aged 39 to 73 years) with valid BMD estimated by quantitative ultrasound of the heel (eBMD) at baseline were included. Linear regression analyses and inverse-variance weighted method were used as main methods for observational and one-sample MR analyses, respectively, to investigate the associations between self-reported sleep traits (sleep duration, chronotype, daytime sleepiness, and insomnia) and low eBMD. Furthermore, sensitivity analyses were performed in subgroups based on sex and menopause in both observational and MR analyses. In observational analyses, short/long sleep, insomnia, and definite eveningness were associated with low eBMD (short sleep: ß = -0.045, effect in standard deviation change of rank-based inverse normally transformed eBMD; long sleep: ß = -0.028; sometimes insomnia: ß = -0.012; usually insomnia: ß = -0.021; definite eveningness: ß = -0.047), whereas definite morningness was associated with decreased risk of low eBMD (ß = 0.011). Subgroup analyses suggested associations of short/long sleep and definite eveningness with low eBMD among men, short sleep with low eBMD among premenopausal women, and short sleep, eveningness, and daytime sleepiness among postmenopausal women. In bidirectional MR analyses, there was no causal relationship between sleep traits and eBMD in either overall sample or subgroup analyses. In summary, although observational analysis showed a robust association of low eBMD with sleep duration, chronotype, and insomnia, there was no evidence of causal relationship as suggested by MR analysis. © 2021 American Society for Bone and Mineral Research (ASBMR).

A framework to decipher the genetic architecture of combinations of complex diseases: applications in cardiovascular medicine.

MOTIVATION: Currently, most genome-wide association studies (GWAS) are studies of a single disease against controls. However, an individual is often affected by more than one condition. For example, coronary artery disease (CAD) is often comorbid with type 2 diabetes mellitus (T2DM). Similarly, it is clinically meaningful to study patients with one disease but without a related comorbidity. For example, obese T2DM may have different pathophysiology from nonobese T2DM. RESULTS: We developed a statistical framework (CombGWAS) to uncover susceptibility variants for comorbid disorders (or a disorder without comorbidity), using GWAS summary statistics only. In essence, we mimicked a case-control GWAS in which the cases are affected with comorbidities or a disease without comorbidity. We extended our methodology to analyze continuous traits with clinically meaningful categories (e.g. lipids), and combination of more than two traits. We verified the feasibility and validity of our method by applying it to simulated scenarios and four cardiometabolic (CM) traits. In total, we identified 384 and 587 genomic risk loci respectively for 6 comorbidities and 12 CM disease 'subtypes' without a relevant comorbidity. Genetic correlation analysis revealed that some subtypes may be biologically distinct from others. Further Mendelian randomization analysis showed differential causal effects of different subtypes to relevant complications. For example, we found that obese T2DM is causally related to increased risk of CAD (P = 2.62E-11). AVAILABILITY AND IMPLEMENTATION: R code is available at: https://github.com/LiangyingYin/CombGWAS. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.