Search

Abstract The spatial arrangement of immune cells within the tumor microenvironment (TME) and their interactions play critical roles in the initiation and development of cancer. Several advanced technologies such as imaging mass cytometry (IMC) providing the immunological landscape of the TME with single-cell resolution. In this study, we develop a new method to quantify the spatial proximity between different cell types based on single-cell spatial data. Using this method on IMC data from 416 lung adenocarcinoma patients, we show that the proximity between different cell types is more correlated with patient prognosis compared to the traditional features such immune cell density and fractions. Consistent with previous reports, our results validate that proximity of T helper (Th) and B cells to cancer cells is associated with survival benefits. More importantly, we discover that the proximity of M2 macrophages to multiple immune cells is associated with poor prognosis. When Th/B cells are stratified into M2-distal and M2-proximal, the abundance of the former but not the latter category of Th/B cells is correlated with enhanced patient survival. Additionally, the abundance of M2-distal and M2-proximal cytotoxic T cells (Tc) is respectively associated with good and poor prognosis. Our results indicate that the prognostic effect of Th, Tc, and B cells in the tumor microenvironment is modulated by the nearby M2 macrophages. The proposed new method proposed can be readily applied to all single-cell spatial data for revealing functional impact of immune cell interactions.

Abstract Background Although polygenic risk score (PRS) has emerged as a promising tool for predicting cancer risk from genome-wide association studies (GWAS), the individual-level accuracy of lung cancer PRS and the extent to which its impact on subsequent clinical applications remains largely unexplored. Methods Lung cancer PRSs and confidence/credible interval (CI) were constructed using two statistical approaches for each individual: (1) the weighted sum of 16 GWAS-derived significant SNP loci and the CI through the bootstrapping method (PRS-16-CV) and (2) LDpred2 and the CI through posteriors sampling (PRS-Bayes), among 17,166 lung cancer cases and 12,894 controls with European ancestry from the International Lung Cancer Consortium. Individuals were classified into different genetic risk subgroups based on the relationship between their own PRS mean/PRS CI and the population level threshold. Results Considerable variances in PRS point estimates at the individual level were observed for both methods, with an average standard deviation (s.d.) of 0.12 for PRS-16-CV and a much larger s.d. of 0.88 for PRS-Bayes. Using PRS-16-CV, only 25.0% of individuals with PRS point estimates in the lowest decile of PRS and 16.8% in the highest decile have their entire 95% CI fully contained in the lowest and highest decile, respectively, while PRS-Bayes was unable to find any eligible individuals. Only 19% of the individuals were concordantly identified as having high genetic risk (> 90th percentile) using the two PRS estimators. An increased relative risk of lung cancer comparing the highest PRS percentile to the lowest was observed when taking the CI into account (OR = 2.73, 95% CI: 2.12–3.50, P-value = 4.13 × 10−15) compared to using PRS-16-CV mean (OR = 2.23, 95% CI: 1.99–2.49, P-value = 5.70 × 10−46). Improved risk prediction performance with higher AUC was consistently observed in individuals identified by PRS-16-CV CI, and the best performance was achieved by incorporating age, gender, and detailed smoking pack-years (AUC: 0.73, 95% CI = 0.72–0.74). Conclusions Lung cancer PRS estimates using different methods have modest correlations at the individual level, highlighting the importance of considering individual-level uncertainty when evaluating the practical utility of PRS.

Abstract Background Expression quantitative trait loci (eQTLs) analyses have been widely used to identify genetic variants associated with gene expression levels to understand what molecular mechanisms underlie genetic traits. The resultant eQTLs might affect the expression of associated genes through transcriptional or post-transcriptional regulation. In this study, we attempt to distinguish these two types of regulation by identifying genetic variants associated with mRNA stability of genes (stQTLs). Results Here, we presented a computational framework that takes advantage of recently developed methods to infer the mRNA stability of genes based on RNA-seq data and performed association analysis to identify stQTLs. Using the Genotype-Tissue Expression (GTEx) lung RNA-Seq data, we identified a total of 142,801 stQTLs for 3942 genes and 186,132 eQTLs for 4751 genes from 15,122,700 genetic variants for 13,476 genes on the autosomes, respectively. Interestingly, our results indicated that stQTLs were enriched in the CDS and 3’UTR regions, while eQTLs are enriched in the CDS, 3’UTR, 5’UTR, and upstream regions. We also found that stQTLs are more likely than eQTLs to overlap with RNA binding protein (RBP) and microRNA (miRNA) binding sites. Our analyses demonstrate that simultaneous identification of stQTLs and eQTLs can provide more mechanistic insight on the association between genetic variants and gene expression levels.

Abstract Background Epidemiological and experimental evidence has linked chronic inflammation to cancer aetiology. It is unclear whether associations for specific inflammatory biomarkers are causal or due to bias. In order to examine whether altered genetically predicted concentration of circulating cytokines are associated with cancer development, we performed a two-sample Mendelian randomisation (MR) analysis. Methods Up to 31,112 individuals of European descent were included in genome-wide association study (GWAS) meta-analyses of 47 circulating cytokines. Single nucleotide polymorphisms (SNPs) robustly associated with the cytokines, located in or close to their coding gene (cis), were used as instrumental variables. Inverse-variance weighted MR was used as the primary analysis, and the MR assumptions were evaluated in sensitivity and colocalization analyses and a false discovery rate (FDR) correction for multiple comparisons was applied. Corresponding germline GWAS summary data for five cancer outcomes (breast, endometrial, lung, ovarian, and prostate), and their subtypes were selected from the largest cancer-specific GWASs available (cases ranging from 12,906 for endometrial to 133,384 for breast cancer). Results There was evidence of inverse associations of macrophage migration inhibitory factor with breast cancer (OR per SD = 0.88, 95% CI 0.83 to 0.94), interleukin-1 receptor antagonist with endometrial cancer (0.86, 0.80 to 0.93), interleukin-18 with lung cancer (0.87, 0.81 to 0.93), and beta-chemokine-RANTES with ovarian cancer (0.70, 0.57 to 0.85) and positive associations of monokine induced by gamma interferon with endometrial cancer (3.73, 1.86 to 7.47) and cutaneous T-cell attracting chemokine with lung cancer (1.51, 1.22 to 1.87). These associations were similar in sensitivity analyses and supported in colocalization analyses. Conclusions Our study adds to current knowledge on the role of specific inflammatory biomarker pathways in cancer aetiology. Further validation is needed to assess the potential of these cytokines as pharmacological or lifestyle targets for cancer prevention.

Abstract Background Lung adenocarcinoma, the most common type of lung cancer, has a high level of morphologic heterogeneity and is composed of tumor cells of multiple histological subtypes. It has been reported that immune cell infiltration significantly impacts clinical outcomes of patients with lung adenocarcinoma. However, it is unclear whether histologic subtyping can reflect the tumor immune microenvironment, and whether histologic subtyping can be applied for therapeutic stratification of the current standard of care. Methods We inferred immune cell infiltration levels using a histological subtype-specific gene expression dataset. From differential gene expression analysis between different histological subtypes, we developed two gene signatures to computationally determine the relative abundance of lepidic and solid components (denoted as the L-score and S-score, respectively) in lung adenocarcinoma samples. These signatures enabled us to investigate the relationship between histological composition and clinical outcomes in lung adenocarcinoma using previously published datasets. Results We found dramatic immunological differences among histological subtypes. Differential gene expression analysis showed that the lepidic and solid subtypes could be differentiated based on their gene expression patterns while the other subtypes shared similar gene expression patterns. Our results indicated that higher L-scores were associated with prolonged survival, and higher S-scores were associated with shortened survival. L-scores and S-scores were also correlated with global genomic features such as tumor mutation burdens and driver genomic events. Interestingly, we observed significantly decreased L-scores and increased S-scores in lung adenocarcinoma samples with EGFR gene amplification but not in samples with EGFR gene mutations. In lung cancer cell lines, we observed significant correlations between L-scores and cell sensitivity to a number of targeted drugs including EGFR inhibitors. Moreover, lung cancer patients with higher L-scores were more likely to benefit from immune checkpoint blockade therapy. Conclusions Our findings provided further insights into evaluating histology composition in lung adenocarcinoma. The established signatures reflected that lepidic and solid subtypes in lung adenocarcinoma would be associated with prognosis, genomic features, and responses to targeted therapy and immunotherapy. The signatures therefore suggested potential clinical translation in predicting patient survival and treatment responses. In addition, our framework can be applied to other types of cancer with heterogeneous histological subtypes.

Abstract Background Aberrant Wnt signalling, regulating cell development and stemness, influences the development of many cancer types. The Aryl hydrocarbon receptor (AhR) mediates tumorigenesis of environmental pollutants. Complex interaction patterns of genes assigned to AhR/Wnt-signalling were recently associated with lung cancer susceptibility. Aim To assess the association and predictive ability of AhR/Wnt-genes with lung cancer in cases and controls of European descent. Methods Odds ratios (OR) were estimated for genomic variants assigned to the Wnt agonist and the antagonistic genes DKK2, DKK3, DKK4, FRZB, SFRP4 and Axin2. Logistic regression models with variable selection were trained, validated and tested to predict lung cancer, at which other previously identified SNPs that have been robustly associated with lung cancer risk could also enter the model. Furthermore, decision trees were created to investigate variant × variant interaction. All analyses were performed for overall lung cancer and for subgroups. Results No genome-wide significant association of AhR/Wnt-genes with overall lung cancer was observed, but within the subgroups of ever smokers (e.g., maker rs2722278 SFRP4; OR = 1.20; 95% CI 1.13–1.27; p = 5.6 × 10–10) and never smokers (e.g., maker rs1133683 Axin2; OR = 1.27; 95% CI 1.19–1.35; p = 1.0 × 10–12). Although predictability is poor, AhR/Wnt-variants are unexpectedly overrepresented in optimized prediction scores for overall lung cancer and for small cell lung cancer. Remarkably, the score for never-smokers contained solely two AhR/Wnt-variants. The optimal decision tree for never smokers consists of 7 AhR/Wnt-variants and only two lung cancer variants. Conclusions The role of variants belonging to Wnt/AhR-pathways in lung cancer susceptibility may be underrated in main-effects association analysis. Complex interaction patterns in individuals of European descent have moderate predictive capacity for lung cancer or subgroups thereof, especially in never smokers.

Abstract Background Over the past decades, approaches for diagnosing and treating cancer have seen significant improvement. However, the variability of patient and tumor characteristics has limited progress on methods for prognosis prediction. The development of high-throughput omics technologies now provides multiple approaches for characterizing tumors. Although a large number of published studies have focused on integration of multi-omics data and use of pathway-level models for cancer prognosis prediction, there still exists a gap of knowledge regarding the prognostic landscape across multi-omics data for multiple cancer types using both gene-level and pathway-level predictors. Methods In this study, we systematically evaluated three often available types of omics data (gene expression, copy number variation and somatic point mutation) covering both DNA-level and RNA-level features. We evaluated the landscape of predictive performance of these three omics modalities for 33 cancer types in the TCGA using a Lasso or Group Lasso-penalized Cox model and either gene or pathway level predictors. Results We constructed the prognostic landscape using three types of omics data for 33 cancer types on both the gene and pathway levels. Based on this landscape, we found that predictive performance is cancer type dependent and we also highlighted the cancer types and omics modalities that support the most accurate prognostic models. In general, models estimated on gene expression data provide the best predictive performance on either gene or pathway level and adding copy number variation or somatic point mutation data to gene expression data does not improve predictive performance, with some exceptional cohorts including low grade glioma and thyroid cancer. In general, pathway-level models have better interpretative performance, higher stability and smaller model size across multiple cancer types and omics data types relative to gene-level models. Conclusions Based on this landscape and comprehensively comparison, models estimated on gene expression data provide the best predictive performance on either gene or pathway level. Pathway-level models have better interpretative performance, higher stability and smaller model size relative to gene-level models.

Abstract Background Genome-wide association studies (GWAS) have proven successful in predicting genetic risk of disease using single-locus models; however, identifying single nucleotide polymorphism (SNP) interactions at the genome-wide scale is limited due to computational and statistical challenges. We addressed the computational burden encountered when detecting SNP interactions for survival analysis, such as age of disease-onset. To confront this problem, we developed a novel algorithm, called the Efficient Survival Multifactor Dimensionality Reduction (ES-MDR) method, which used Martingale Residuals as the outcome parameter to estimate survival outcomes, and implemented the Quantitative Multifactor Dimensionality Reduction method to identify significant interactions associated with age of disease-onset. Methods To demonstrate efficacy, we evaluated this method on two simulation data sets to estimate the type I error rate and power. Simulations showed that ES-MDR identified interactions using less computational workload and allowed for adjustment of covariates. We applied ES-MDR on the OncoArray-TRICL Consortium data with 14,935 cases and 12,787 controls for lung cancer (SNPs = 108,254) to search over all two-way interactions to identify genetic interactions associated with lung cancer age-of-onset. We tested the best model in an independent data set from the OncoArray-TRICL data. Results Our experiment on the OncoArray-TRICL data identified many one-way and two-way models with a single-base deletion in the noncoding region of BRCA1 (HR 1.24, P = 3.15 × 10–15), as the top marker to predict age of lung cancer onset. Conclusions From the results of our extensive simulations and analysis of a large GWAS study, we demonstrated that our method is an efficient algorithm that identified genetic interactions to include in our models to predict survival outcomes.

Abstract Background Genomic profiling of solid human tumors by projects such as The Cancer Genome Atlas (TCGA) has provided important information regarding the somatic alterations that drive cancer progression and patient survival. Although researchers have successfully leveraged TCGA data to build prognostic models, most efforts have focused on specific cancer types and a targeted set of gene-level predictors. Less is known about the prognostic ability of pathway-level variables in a pan-cancer setting. To address these limitations, we systematically evaluated and compared the prognostic ability of somatic point mutation (SPM) and copy number variation (CNV) data, gene-level and pathway-level models for a diverse set of TCGA cancer types and predictive modeling approaches. Results We evaluated gene-level and pathway-level penalized Cox proportional hazards models using SPM and CNV data for 29 different TCGA cohorts. We measured predictive accuracy as the concordance index for predicting survival outcomes. Our comprehensive analysis suggests that the use of pathway-level predictors did not offer superior predictive power relative to gene-level models for all cancer types but had the advantages of robustness and parsimony. We identified a set of cohorts for which somatic alterations could not predict prognosis, and a unique cohort LGG, for which SPM data was more predictive than CNV data and the predictive accuracy is good for all model types. We found that the pathway-level predictors provide superior interpretative value and that there is often a serious collinearity issue for the gene-level models while pathway-level models avoided this issue. Conclusion Our comprehensive analysis suggests that when using somatic alterations data for cancer prognosis prediction, pathway-level models are more interpretable, stable and parsimonious compared to gene-level models. Pathway-level models also avoid the issue of collinearity, which can be serious for gene-level somatic alterations. The prognostic power of somatic alterations is highly variable across different cancer types and we have identified a set of cohorts for which somatic alterations could not predict prognosis. In general, CNV data predicts prognosis better than SPM data with the exception of the LGG cohort.

Abstract Background Cancer prognosis prediction is valuable for patients and clinicians because it allows them to appropriately manage care. A promising direction for improving the performance and interpretation of expression-based predictive models involves the aggregation of gene-level data into biological pathways. While many studies have used pathway-level predictors for cancer survival analysis, a comprehensive comparison of pathway-level and gene-level prognostic models has not been performed. To address this gap, we characterized the performance of penalized Cox proportional hazard models built using either pathway- or gene-level predictors for the cancers profiled in The Cancer Genome Atlas (TCGA) and pathways from the Molecular Signatures Database (MSigDB). Results When analyzing TCGA data, we found that pathway-level models are more parsimonious, more robust, more computationally efficient and easier to interpret than gene-level models with similar predictive performance. For example, both pathway-level and gene-level models have an average Cox concordance index of ~ 0.85 for the TCGA glioma cohort, however, the gene-level model has twice as many predictors on average, the predictor composition is less stable across cross-validation folds and estimation takes 40 times as long as compared to the pathway-level model. When the complex correlation structure of the data is broken by permutation, the pathway-level model has greater predictive performance while still retaining superior interpretative power, robustness, parsimony and computational efficiency relative to the gene-level models. For example, the average concordance index of the pathway-level model increases to 0.88 while the gene-level model falls to 0.56 for the TCGA glioma cohort using survival times simulated from uncorrelated gene expression data. Conclusion The results of this study show that when the correlations among gene expression values are low, pathway-level analyses can yield better predictive performance, greater interpretative power, more robust models and less computational cost relative to a gene-level model. When correlations among genes are high, a pathway-level analysis provides equivalent predictive power compared to a gene-level analysis while retaining the advantages of interpretability, robustness and computational efficiency.

Abstract Background Cancer develops when pathways controlling cell survival, cell fate or genome maintenance are disrupted by the somatic alteration of key driver genes. Understanding how pathway disruption is driven by somatic alterations is thus essential for an accurate characterization of cancer biology and identification of therapeutic targets. Unfortunately, current cancer pathway analysis methods fail to fully model the relationship between somatic alterations and pathway activity. Results To address these limitations, we developed a multi-omics method for identifying biologically important pathways and genes in human cancer. Our approach combines single-sample pathway analysis with multi-stage, lasso-penalized regression to find pathways whose gene expression can be explained largely in terms of gene-level somatic alterations in the tumor. Importantly, this method can analyze case-only data sets, does not require information regarding pathway topology and supports personalized pathway analysis using just somatic alteration data for a limited number of cancer-associated genes. The practical effectiveness of this technique is illustrated through an analysis of data from The Cancer Genome Atlas using gene sets from the Molecular Signatures Database. Conclusions Novel insights into the pathophysiology of human cancer can be obtained from statistical models that predict expression-based pathway activity in terms of non-silent somatic mutations and copy number variation. These models enable the identification of biologically important pathways and genes and support personalized pathway analysis in cases where gene expression data is unavailable.

Abstract Background Because driver mutations provide selective advantage to the mutant clone, they tend to occur at a higher frequency in tumor samples compared to selectively neutral (passenger) mutations. However, mutation frequency alone is insufficient to identify cancer genes because mutability is influenced by many gene characteristics, such as size, nucleotide composition, etc. The goal of this study was to identify gene characteristics associated with the frequency of somatic mutations in the gene in tumor samples. Results We used data on somatic mutations detected by genome wide screens from the Catalog of Somatic Mutations in Cancer (COSMIC). Gene size, nucleotide composition, expression level of the gene, relative replication time in the cell cycle, level of evolutionary conservation and other gene characteristics (totaling 11) were used as predictors of the number of somatic mutations. We applied stepwise multiple linear regression to predict the number of mutations per gene. Because missense, nonsense, and frameshift mutations are associated with different sets of gene characteristics, they were modeled separately. Gene characteristics explain 88% of the variation in the number of missense, 40% of nonsense, and 23% of frameshift mutations. Comparisons of the observed and expected numbers of mutations identified genes with a higher than expected number of mutations– positive outliers. Many of these are known driver genes. A number of novel candidate driver genes was also identified. Conclusions By comparing the observed and predicted number of mutations in a gene, we have identified known cancer-associated genes as well as 111 novel cancer associated genes. We also showed that adding the number of silent mutations per gene reported by genome/exome wide screens across all cancer type (COSMIC data) as a predictor substantially exceeds predicting accuracy of the most popular cancer gene predicting tool - MutsigCV.

Abstract Background Accurate inference of genetic ancestry is of fundamental interest to many biomedical, forensic, and anthropological research areas. Genetic ancestry memberships may relate to genetic disease risks. In a genome association study, failing to account for differences in genetic ancestry between cases and controls may also lead to false-positive results. Although a number of strategies for inferring and taking into account the confounding effects of genetic ancestry are available, applying them to large studies (tens thousands samples) is challenging. The goal of this study is to develop an approach for inferring genetic ancestry of samples with unknown ancestry among closely related populations and to provide accurate estimates of ancestry for application to large-scale studies. Methods In this study we developed a novel distance-based approach, Ancestry Inference using Principal component analysis and Spatial analysis (AIPS) that incorporates an Inverse Distance Weighted (IDW) interpolation method from spatial analysis to assign individuals to population memberships. Results We demonstrate the benefits of AIPS in analyzing population substructure, specifically related to the four most commonly used tools EIGENSTRAT, STRUCTURE, fastSTRUCTURE, and ADMIXTURE using genotype data from various intra-European panels and European-Americans. While the aforementioned commonly used tools performed poorly in inferring ancestry from a large number of subpopulations, AIPS accurately distinguished variations between and within subpopulations. Conclusions Our results show that AIPS can be applied to large-scale data sets to discriminate the modest variability among intra-continental populations as well as for characterizing inter-continental variation. The method we developed will protect against spurious associations when mapping the genetic basis of a disease. Our approach is more accurate and computationally efficient method for inferring genetic ancestry in the large-scale genetic studies.

Abstract Background Nearly 6 million deaths and over a half trillion dollars in healthcare costs worldwide are attributed to tobacco smoking each year. Extensive research efforts have been pursued to elucidate the molecular underpinnings of smoking addiction and facilitate cessation. In this study, we genotyped and obtained both resting state and task-based functional magnetic resonance imaging from 64 non-smokers and 42 smokers. Smokers were imaged after having smoked normally (“sated”) and after having not smoked for at least 12 h (“abstinent”). Results While abstinent smokers did not differ from non-smokers with respect to pairwise resting state functional connectivities (RSFCs) between 12 brain regions of interest, RSFCs involving the caudate and putamen of sated smokers significantly differed from those of non-smokers (P