ARTEMIS: An independently validated prognostic prediction model of breast cancer incorporating epigenetic biomarkers with main effects and gene-gene interactions.

INTRODUCTION: Breast cancer, a heterogeneous disease, is influenced by multiple genetic and epigenetic factors. The majority of prognostic models for breast cancer focus merely on the main effects of predictors, disregarding the crucial impacts of gene-gene interactions on prognosis. OBJECTIVES: Using DNA methylation data derived from nine independent breast cancer cohorts, we developed an independently validated prognostic prediction model of breast cancer incorporating epigenetic biomarkers with main effects and gene-gene interactions (ARTEMIS) with an innovative 3-D modeling strategy. ARTEMIS was evaluated for discrimination ability using area under the receiver operating characteristics curve (AUC), and calibration using expected and observed (E/O) ratio. Additionally, we conducted decision curve analysis to evaluate its clinical efficacy by net benefit (NB) and net reduction (NR). Furthermore, we conducted a systematic review to compare its performance with existing models. RESULTS: ARTEMIS exhibited excellent risk stratification ability in identifying patients at high risk of mortality. Compared to those below the 25th percentile of ARTEMIS scores, patients with above the 90th percentile had significantly lower overall survival time (HR = 15.43, 95% CI: 9.57-24.88, P = 3.06 × 10-29). ARTEMIS demonstrated satisfactory discrimination ability across four independent populations, with pooled AUC3-year = 0.844 (95% CI: 0.805-0.883), AUC5-year = 0.816 (95% CI: 0.775-0.857), and C-index = 0.803 (95% CI: 0.776-0.830). Meanwhile, ARTEMIS had well calibration performance with pooled E/O ratio 1.060 (95% CI: 1.038-1.083) and 1.090 (95% CI: 1.057-1.122) for 3- and 5-year survival prediction, respectively. Additionally, ARTEMIS is a clinical instrument with acceptable cost-effectiveness for detecting breast cancer patients at high risk of mortality (Pt = 0.4: NB3-year = 19‰, NB5-year = 62‰; NR3-year = 69.21%, NR5-year = 56.01%). ARTEMIS has superior performance compared to existing models in terms of accuracy, extrapolation, and sample size, as indicated by the systematic review. ARTEMIS is implemented as an interactive online tool available at http://bigdata.njmu.edu.cn/ARTEMIS/. CONCLUSION: ARTEMIS is an efficient and practical tool for breast cancer prognostic prediction.

Spicy food consumption and risk of vascular disease: Evidence from a large-scale Chinese prospective cohort of 0.5 million people.

BACKGROUND: Spicy food consumption has been reported to be inversely associated with mortality from multiple diseases. However, the effect of spicy food intake on the incidence of vascular diseases in the Chinese population remains unclear. This study was conducted to explore this association. METHODS: This study was performed using the large-scale China Kadoorie Biobank (CKB) prospective cohort of 486,335 participants. The primary outcomes were vascular disease, ischemic heart disease (IHD), major coronary events (MCEs), cerebrovascular disease, stroke, and non-stroke cerebrovascular disease. A Cox proportional hazards regression model was used to assess the association between spicy food consumption and incident vascular diseases. Subgroup analysis was also performed to evaluate the heterogeneity of the association between spicy food consumption and the risk of vascular disease stratified by several basic characteristics. In addition, the joint effects of spicy food consumption and the healthy lifestyle score on the risk of vascular disease were also evaluated, and sensitivity analyses were performed to assess the reliability of the association results. RESULTS: During a median follow-up time of 12.1 years, a total of 136,125 patients with vascular disease, 46,689 patients with IHD, 10,097 patients with MCEs, 80,114 patients with cerebrovascular disease, 56,726 patients with stroke, and 40,098 patients with non-stroke cerebrovascular disease were identified. Participants who consumed spicy food 1-2 days/week (hazard ratio [HR] = 0.95, 95% confidence interval [95% CI] = [0.93, 0.97], P <0.001), 3-5 days/week (HR = 0.96, 95% CI = [0.94, 0.99], P = 0.003), and 6-7 days/week (HR = 0.97, 95% CI = [0.95, 0.99], P = 0.002) had a significantly lower risk of vascular disease than those who consumed spicy food less than once a week (Ptrend <0.001), especially in those who were younger and living in rural areas. Notably, the disease-based subgroup analysis indicated that the inverse associations remained in IHD (Ptrend = 0.011) and MCEs (Ptrend = 0.002) risk. Intriguingly, there was an interaction effect between spicy food consumption and the healthy lifestyle score on the risk of IHD (Pinteraction = 0.037). CONCLUSIONS: Our findings support an inverse association between spicy food consumption and vascular disease in the Chinese population, which may provide additional dietary guidance for the prevention of vascular diseases.

BHAFT: Bayesian heredity-constrained accelerated failure time models for detecting gene-environment interactions in survival analysis.

In addition to considering the main effects, understanding gene-environment (G × E) interactions is imperative for determining the etiology of diseases and the factors that affect their prognosis. In the existing statistical framework for censored survival outcomes, there are several challenges in detecting G × E interactions, such as handling high-dimensional omics data, diverse environmental factors, and algorithmic complications in survival analysis. The effect heredity principle has widely been used in studies involving interaction identification because it incorporates the dependence of the main and interaction effects. However, Bayesian survival models that incorporate the assumption of this principle have not been developed. Therefore, we propose Bayesian heredity-constrained accelerated failure time (BHAFT) models for identifying main and interaction (M-I) effects with novel spike-and-slab or regularized horseshoe priors to incorporate the assumption of effect heredity principle. The R package rstan was used to fit the proposed models. Extensive simulations demonstrated that BHAFT models had outperformed other existing models in terms of signal identification, coefficient estimation, and prognosis prediction. Biologically plausible G × E interactions associated with the prognosis of lung adenocarcinoma were identified using our proposed model. Notably, BHAFT models incorporating the effect heredity principle could identify both main and interaction effects, which are highly useful in exploring G × E interactions in high-dimensional survival analysis. The code and data used in our paper are available at https://github.com/SunNa-bayesian/BHAFT.

Serial platelet count as a dynamic prediction marker of hospital mortality among septic patients.

Background: Platelets play a critical role in hemostasis and inflammatory diseases. Low platelet count and activity have been reported to be associated with unfavorable prognosis. This study aims to explore the relationship between dynamics in platelet count and in-hospital morality among septic patients and to provide real-time updates on mortality risk to achieve dynamic prediction. Methods: We conducted a multi-cohort, retrospective, observational study that encompasses data on septic patients in the eICU Collaborative Research Database (eICU-CRD) and the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. The joint latent class model (JLCM) was utilized to identify heterogenous platelet count trajectories over time among septic patients. We assessed the association between different trajectory patterns and 28-day in-hospital mortality using a piecewise Cox hazard model within each trajectory. We evaluated the performance of our dynamic prediction model through area under the receiver operating characteristic curve, concordance index (C-index), accuracy, sensitivity, and specificity calculated at predefined time points. Results: Four subgroups of platelet count trajectories were identified that correspond to distinct in-hospital mortality risk. Including platelet count did not significantly enhance prediction accuracy at early stages (day 1 C-indexDynamic  vs C-indexWeibull: 0.713 vs 0.714). However, our model showed superior performance to the static survival model over time (day 14 C-indexDynamic  vs C-indexWeibull: 0.644 vs 0.617). Conclusions: For septic patients in an intensive care unit, the rapid decline in platelet counts is a critical prognostic factor, and serial platelet measures are associated with prognosis.

Elevated remnant cholesterol is a risk factor for acute ischemic stroke.

OBJECTIVES: Remnant cholesterol (RC) is thought to be an important pathogenic risk factor for atherosclerosis, however, the relationship between RC and acute ischemic stroke (AIS) is still unclear. This study aimed to determine whether fasting blood RC level is an independent risk factor for AIS. MATERIALS AND METHODS: A retrospective analysis was performed on 650 patients with AIS and 598 healthy controls during the same time period. The association between RC and AIS was investigated using binary logistic regression, and the relationship between RC and AIS risk was demonstrated using Restricted Cubic Splines (RCS). RESULTS: RC was significantly higher in the AIS group compared with control group, and was an independent risk factor for AIS when the covariates were not adjusted;After adjusting some covariates, RC was still an independent risk factor for AIS. The RCS analysis found the risk was non-linear: when RC concentration was less than 0.69 mol/L, the risk of AIS increased with the elevation of RC, and when RC concentration was more than or equal to 0.69 mol/L, the risk of AIS was insignificant with the elevation of RC. Correlation analysis revealed that RC was associated with diabetes and fasting glucose. Further analysis revealed that the incidence of AIS in diabetic patients increased significantly with the increase of RC, and RCS analysis revealed that the risk of AIS in diabetic patients increased with the increase of RC when RC was more than 1.15 mol/L. CONCLUSIONS: This study confirms RC as an independent risk factor for AIS, which highlights a distinct non-linear association between RC levels and AIS risk. These findings suggest the need for targeted AIS risk assessment strategies, especially in diabetic patients, and underscore the relevance of RC as a biomarker in AIS risk stratification.

A cross-trait study of lung cancer and its related respiratory diseases based on large-scale exome sequencing population.

Background: Genome-wide association studies (GWASs) explain the genetic susceptibility between diseases and common variants. Nevertheless, with the appearance of large-scale sequencing profiles, we could explore the rare coding variants in disease pathogenesis. Methods: We estimated the genetic correlation of nine respiratory diseases and lung cancer in the UK Biobank (UKB) by linkage disequilibrium score regression (LDSC). Then, we performed exome-wide association studies at single-variant level and gene-level for lung cancer and lung cancer-related respiratory diseases using the whole-exome sequencing (WES) data of 427,934 European participants. Cross-trait meta-analysis was conducted by association analysis based on subsets (ASSET) to identify the pleiotropic variants, while in-silico functional analysis was performed to explore their function. Causal mediation analysis was used to explore whether these pleiotropic variants lead to lung cancer is mediated by affecting the chronic respiratory diseases. Results: Five respiratory diseases [emphysema, pneumonia, asthma, chronic obstructive pulmonary disease (COPD), and fibrosis] were genetically correlated with lung cancer. We identified 102 significant independent variants at single-variant levels for lung cancer and five lung cancer-related diseases. 15:78590583:G>A (missense variant in CHRNA5) was shared in lung cancer, emphysema, and COPD. Meanwhile, 14 significant genes and 87 suggestive genes were identified in gene-based association tests, including HSD3B7 (lung cancer), SRSF2 (pneumonia), TNXB (asthma), TERT (fibrosis), MOSPD3 (emphysema). Based on the cross-trait meta-analysis, we detected 145 independent pleiotropic variants. We further identified abundant pathways with significant enrichment effects, demonstrating that these pleiotropic genes were functional. Meanwhile, the proportion of mediation effects of these variants ranged from 6 to 23 (emphysema: 23%; COPD: 20%; pneumonia: 20%; fibrosis: 7%; asthma: 6%) through these five respiratory diseases to the incidence of lung cancer. Conclusions: The identified shared genetic variants, genes, biological pathways, and potential intermediate causal pathways provide a basis for further exploration of the relationship between lung cancer and respiratory diseases.

Proteome-wide Mendelian randomization identifies causal plasma proteins in lung cancer.

Plasma proteins are promising biomarkers and potential drug targets in lung cancer. To evaluate the causal association between plasma proteins and lung cancer, we performed proteome-wide Mendelian randomization meta-analysis (PW-MR-meta) based on lung cancer genome-wide association studies (GWASs), protein quantitative trait loci (pQTLs) of 4,719 plasma proteins in deCODE and 4,775 in Fenland. Further, causal-protein risk score (CPRS) was developed based on causal proteins and validated in the UK Biobank. 270 plasma proteins were identified using PW-MR meta-analysis, including 39 robust causal proteins (both FDR-q < 0.05) and 78 moderate causal proteins (FDR-q < 0.05 in one and p < 0.05 in another). The CPRS had satisfactory performance in risk stratification for lung cancer (top 10% CPRS:Hazard ratio (HR) (95%CI):4.33(2.65-7.06)). The CPRS [AUC (95%CI): 65.93 (62.91-68.78)] outperformed the traditional polygenic risk score (PRS) [AUC (95%CI): 55.71(52.67-58.59)]. Our findings offer further insight into the genetic architecture of plasma proteins for lung cancer susceptibility.

Identification of differentially expressed exosome proteins in serum as potential biomarkers for cognitive impairments in cerebral small vessel disease.

BACKGROUND: Cognitive impairment arising from cerebral small vessel disease (CSVD) represents a critical subtype of vascular cognitive impairments (VCI) and is the primary cause of vascular dementia. However, identifying reliable clinical and laboratory indicators for this disease remain elusive. We hypothesize that plasma exosome proteins hold the potential to serve as biomarkers for the onset of cognitive dysfunction associated with cerebrovascular diseases. METHODS: We employed TMT-based proteomics to discern variations in serum exosome proteomes between individuals with cognitive impairments due to CSVD and healthy volunteers. RESULTS: Each group comprised 18 subjects, and through differential expression analysis, we identified 22 down-regulated and 8 up-regulated proteins between the two groups. Our research revealed 30 differentially expressed plasma exosome proteins, including histone, proteasome, clusterin and coagulation factor XIII, in individuals with cognitive impairments caused by CSVD. CONCLUSION: The 30 differentially expressed plasma exosome proteins identified in our study are promising as biomarkers for diagnosing cognitive impairments resulting from CSVD. These findings may help us better understand the underlying pathological mechanisms involved in the diseases.

Systematic characterization of m6A proteomics across 12 cancer types: a multi-omics integration study.

The modification patterns of N6-methyladenosine (m6A) regulators and interacting genes are deeply involved in tumors. However, the effect of m6A modification patterns on human proteomics remains largely unknown. We evaluated the molecular characteristics and clinical relevance of m6A modification proteomics patterns among 1013 pan-cancer samples from the Clinical Proteomic Tumor Analysis Consortium (CPTAC). More than half of the m6A proteins were expressed at higher levels in tumor tissues and presented oncogenic characteristics. Furthermore, we performed multi-omics analyses integrating with transcriptomics data of m6A regulators and interactive coding and non-coding RNAs and developed a m6A multi-omics signature to identify potential m6A modification target proteins across global proteomics. It was significantly associated with overall survival in nine cancer types, tumor mutation burden (P = 0.01), and immune checkpoints including PD-L1 (P = 4.9 × 10-8) and PD-1 (P < 0.01). We identified 51 novel proteins associated with the multi-omics signature (PFDR < 0.05). These proteins were functional through pathway enrichment analyses. The protein with the highest hit frequency was CHORDC1, which was significantly up-regulated in tumor tissues in nine cancer types. Its higher abundance was significantly associated with a poorer prognosis in seven cancer types. The identified m6A target proteins might provide infomation for the study of molecular mechanism of cancer.

A cross-disorder study to identify causal relationships, shared genetic variants, and genes across 21 digestive disorders.

Digestive disorders are a significant contributor to the global burden of disease and seriously affect human quality of life. Research has already confirmed the presence of pleiotropic genetic loci among digestive disorders, and studies have explored shared genetic factors among pan-cancers, including various malignant digestive disorders. However, most cross-phenotype studies within the digestive tract system have been limited to a few traits, with no systematic coverage of common benign and malignant digestive disorders. Here, we analyzed data from the UK Biobank to investigate 21 digestive disorders, exploring the genetic correlations and causal relationships between diseases, as well as the common genetic factors and potential biological pathways driving these relationships. Our findings confirmed the extensive genetic correlation and causal relationship between digestive disorders, providing important insights into the genetic etiology, causality, disease prevention, and clinical treatment of diseases.

Genomic analysis of a new heterotic maize group reveals key loci for pedigree breeding.

Genome-wide analyses of maize populations have clarified the genetic basis of crop domestication and improvement. However, limited information is available on how breeding improvement reshaped the genome in the process of the formation of heterotic groups. In this study, we identified a new heterotic group (X group) based on an examination of 512 Chinese maize inbred lines. The X group was clearly distinct from the other non-H&L groups, implying that X × HIL is a new heterotic pattern. We selected the core inbred lines for an analysis of yield-related traits. Almost all yield-related traits were better in the X lines than those in the parental lines, indicating that the primary genetic improvement in the X group during breeding was yield-related traits. We generated whole-genome sequences of these lines with an average coverage of 17.35× to explore genome changes further. We analyzed the identity-by-descent (IBD) segments transferred from the two parents to the X lines and identified 29 and 28 IBD conserved regions (ICRs) from the parents PH4CV and PH6WC, respectively, accounting for 28.8% and 12.8% of the genome. We also identified 103, 89, and 131 selective sweeps (SSWs) using methods that involved the π, Tajima's D, and CLR values, respectively. Notably, 96.13% of the ICRs co-localized with SSWs, indicating that SSW signals concentrated in ICRs. We identified 171 annotated genes associated with yield-related traits in maize both in ICRs and SSWs. To identify the genetic factors associated with yield improvement, we conducted QTL mapping for 240 lines from a DH population (PH4CV × PH6WC, which are the parents of X1132X) for ten key yield-related traits and identified a total of 55 QTLs. Furthermore, we detected three QTL clusters both in ICRs and SSWs. Based on the genetic evidence, we finally identified three key genes contributing to yield improvement in breeding the X group. These findings reveal key loci and genes targeted during pedigree breeding and provide new insights for future genomic breeding.

ATHENA: an independently validated autophagy-related epigenetic prognostic prediction model of head and neck squamous cell carcinoma.

The majority of these existing prognostic models of head and neck squamous cell carcinoma (HNSCC) have unsatisfactory prediction accuracy since they solely utilize demographic and clinical information. Leveraged by autophagy-related epigenetic biomarkers, we aim to develop a better prognostic prediction model of HNSCC incorporating CpG probes with either main effects or gene-gene interactions. Based on DNA methylation data from three independent cohorts, we applied a 3-D analysis strategy to develop An independently validated auTophagy-related epigenetic prognostic prediction model of HEad and Neck squamous cell carcinomA (ATHENA). Compared to prediction models with only demographic and clinical information, ATHENA has substantially improved discriminative ability, prediction accuracy and more clinical net benefits, and shows robustness in different subpopulations, as well as external populations. Besides, epigenetic score of ATHENA is significantly associated with tumor immune microenvironment, tumor-infiltrating immune cell abundances, immune checkpoints, somatic mutation and immunity-related drugs. Taken together these results, ATHENA has the demonstrated feasibility and utility of predicting HNSCC survival ( http://bigdata.njmu.edu.cn/ATHENA/ ).

Trash to treasure: lactate and protein lactylation in maize root impacts response to drought.

Lactate, protein lactylation (Kla), and specifically histone lactylation have recently been shown to regulate antipathogenic immune responses in mammals. Herein, after we confirmed the presence and accumulation of lactate in maize roots under drought conditions, a lactylome profiling analysis revealed that Kla modifications were invariably present in maize roots, that there were obvious differences in the lactylomes of drought-sensitive (B73) vs. drought-tolerant (Jing2416) lines, and that growing Jing2416 under drought conditions caused significant decreases in the lactylation of multiple enzymes responsible for fatty acid degradation. Inspired by findings of histone-Kla based epigenetic regulation of immune functions in animals, we initially discovered 37 Kla sites on 16 histones in the maize genome, and again detected obvious differential histone Kla-mediated trends between two lines by ChIP-Seq. Notably, only 2.7% of genes with differential histone Kla peaks detected during drought stress were commonly present in both lines, a finding demonstrating that abiotic stress triggers distinct epigenetic activities in diverse germplasm while also strongly supporting that a histone Kla layer of regulation is associated with physiological responses to drought stress. Interestingly, exogenous application of spermidine improved the drought tolerance of B73 and substantially altered the levels of lactate, protein lactylation, and histone Kla modification. Thus, beyond extending the known domain of Kla-based biochemical and epigenetic regulation from animal immunity to plant stress physiology, our study suggests the physiological, biochemical, and genetic function of "the best-known metabolic waste", lactate.

WNT9A Affects Late-Onset Acute Respiratory Distress Syndrome and 28-Day Survival: Evidence from a Three-Step Multiomics Study.

Late-onset (more than 48 h after ICU admission) acute respiratory distress syndrome (ARDS) is associated with shorter survival time and higher mortality; however, the underlying molecular targets remain unclear. As the WNT gene family is known to drive inflammation, immunity, and tissue fibrosis, all of which are closely related to the pathogenesis and prognosis of ARDS, we aim to investigate the associations of the WNT family with late-onset ARDS and 28-day survival. Genetic (n = 380), epigenetic (n = 185), transcriptional (n = 160), and protein (n = 300) data of patients with ARDS were extracted from the MEARDS (Molecular Epidemiology of ARDS) cohort. We used sure independence screening to identify late onset-related genetic biomarkers and constructed a genetic score on the basis of eight SNPs, which was associated with risk for late-onset ARDS (odds ratio [OR], 2.72; P = 3.81 × 10-14) and survival (hazard ratio [HR], 1.28; P = 0.008). The associations were further externally validated in the iSPAAR (Identification of SNPs Predisposing to Altered Acute Lung Injury Risk) (ORlate onset, 2.49 [P = 0.006]; HRsurvival, 1.87 [P = 0.045]) and MESSI (Molecular Epidemiology of Severe Sepsis in the ICU) (ORlate onset, 4.12 [P = 0.026]; HRsurvival, 1.45 [P = 0.036]) cohorts. Furthermore, we functionally interrogated the six mapped genes of eight SNPs in the multiomics data and noted associations of WNT9A (WNT family member 9A) in epigenetic (ORlate onset, 2.95 [P = 9.91 × 10-4]; HRsurvival, 1.53 [P = 0.011]) and protein (ORlate onset, 1.42 [P = 0.035]; HRsurvival, 1.38 [P = 0.011]) data. The mediation analysis indicated that the effects of WNT9A on ARDS survival were mediated by late onset (HRindirect, 1.12 [P = 0.014] for genetic data; HRindirect, 1.05 [P = 0.030] for protein data). The essential roles of WNT9A in immunity and fibrosis may explain the different trajectories of recovery and dysfunction between early- and late-onset ARDS, providing clues for ARDS treatment.

Large-scale integration of the non-coding RNAs with DNA methylation in human cancers.

Characterizing influences of DNA methylation (DNAm) on non-coding RNAs (ncRNAs) is important to understand the mechanisms of gene regulation and cancer outcome. In our study, we describe the results of ncRNA quantitative trait methylation sites (ncQTM) analyses on 8,545 samples from The Cancer Genome Atlas (TCGA), 763 samples from the Clinical Proteomic Tumor Analysis Consortium (CPTAC), and 516 samples from Genotype-Tissue Expression (GTEx) to identify the significant associations between DNAm sites and ncRNAs (miRNA, long non-coding RNA [lncRNA], small nuclear RNA [snRNA], small nucleolar RNA [snoRNA], and rRNA) across 32 cancer types. With more than 22 billion tests, we identify 302,764 cis-ncQTMs (6.28% of all tested) and 79,841,728 trans-ncQTMs (1.15% of all tested). Most DNAm sites (70.6% on average) are in trans association, while only 25.2% DNAm sites are in cis association. Further, we develop a subtype named ncmcluster based on cancer-specific ncRNAs thatis associated with tumor microenvironment, clinical outcome, and biological pathways. To comprehensively describe the ncQTM patterns, we developed a database named Pancan-ncQTM (http://bigdata.njmu.edu.cn/Pancan-ncQTM/).

De novo genome assembly and analyses of 12 founder inbred lines provide insights into maize heterosis.

Hybrid maize displays superior heterosis and contributes over 30% of total worldwide cereal production. However, the molecular mechanisms of heterosis remain obscure. Here we show that structural variants (SVs) between the parental lines have a predominant role underpinning maize heterosis. De novo assembly and analyses of 12 maize founder inbred lines (FILs) reveal abundant genetic variations among these FILs and, through expression quantitative trait loci and association analyses, we identify several SVs contributing to genomic and phenotypic differentiations of various heterotic groups. Using a set of 91 diallel-cross F1 hybrids, we found strong positive correlations between better-parent heterosis of the F1 hybrids and the numbers of SVs between the parental lines, providing concrete genomic support for a prevalent role of genetic complementation underlying heterosis. Further, we document evidence that SVs in both ZAR1 and ZmACO2 contribute to yield heterosis in an overdominance fashion. Our results should promote genomics-based breeding of hybrid maize.

OWL: an optimized and independently validated machine learning prediction model for lung cancer screening based on the UK Biobank, PLCO, and NLST populations.

BACKGROUND: A reliable risk prediction model is critically important for identifying individuals with high risk of developing lung cancer as candidates for low-dose chest computed tomography (LDCT) screening. Leveraging a cutting-edge machine learning technique that accommodates a wide list of questionnaire-based predictors, we sought to optimize and validate a lung cancer prediction model. METHODS: We developed an Optimized early Warning model for Lung cancer risk (OWL) using the XGBoost algorithm with 323,344 participants from the England area in UK Biobank (training set), and independently validated it with 93,227 participants from UKB Scotland and Wales area (validation set 1), as well as 70,605 and 66,231 participants in the Prostate, Lung, Colorectal, and Ovarian cancer screening trial (PLCO) control and intervention subpopulations, respectively (validation sets 2 & 3) and 23,138 and 18,669 participants in the United States National Lung Screening Trial (NLST) control and intervention subpopulations, respectively (validation sets 4 & 5). By comparing with three competitive prediction models, i.e., PLCO modified 2012 (PLCOm2012), PLCO modified 2014 (PLCOall2014), and the Liverpool Lung cancer Project risk model version 3 (LLPv3), we assessed the discrimination of OWL by the area under receiver operating characteristic curve (AUC) at the designed time point. We further evaluated the calibration using relative improvement in the ratio of expected to observed lung cancer cases (RIEO), and illustrated the clinical utility by the decision curve analysis. FINDINGS: For general population, with validation set 1, OWL (AUC = 0.855, 95% CI: 0.829-0.880) presented a better discriminative capability than PLCOall2014 (AUC = 0.821, 95% CI: 0.794-0.848) (p < 0.001); with validation sets 2 & 3, AUC of OWL was comparable to PLCOall2014 (AUCPLCOall2014-AUCOWL < 1%). For ever-smokers, OWL outperformed PLCOm2012 and PLCOall2014 among ever-smokers in validation set 1 (AUCOWL = 0.842, 95% CI: 0.814-0.871; AUCPLCOm2012 = 0.792, 95% CI: 0.760-0.823; AUCPLCOall2014 = 0.791, 95% CI: 0.760-0.822, all p < 0.001). OWL remained comparable to PLCOm2012 and PLCOall2014 in discrimination (AUC difference from -0.014 to 0.008) among the ever-smokers in validation sets 2 to 5. In all the validation sets, OWL outperformed LLPv3 among the general population and the ever-smokers. Of note, OWL showed significantly better calibration than PLCOm2012, PLCOall2014 (RIEO from 43.1% to 92.3%, all p < 0.001), and LLPv3 (RIEO from 41.4% to 98.7%, all p < 0.001) in most cases. For clinical utility, OWL exhibited significant improvement in average net benefits (NB) over PLCOall2014 in validation set 1 (NB improvement: 32, p < 0.001); among ever smokers of validation set 1, OWL (average NB = 289) retained significant improvement over PLCOm2012 (average NB = 213) (p < 0.001). OWL had equivalent NBs with PLCOm2012 and PLCOall2014 in PLCO and NLST populations, while outperforming LLPv3 in the three populations. INTERPRETATION: OWL, with a high degree of predictive accuracy and robustness, is a general framework with scientific justifications and clinical utility that can aid in screening individuals with high risks of lung cancer. FUNDING: National Natural Science Foundation of China, the US NIH.