A subregion-based burden test for simultaneous identification of susceptibility loci and subregions within

Whole exome sequencing and machine learning germline analysis of individuals presenting with extreme phenotypes of high and low risk of developing tobacco-associated lung adenocarcinoma

BACKGROUND

Tobacco is the main risk factor for developing lung cancer. Yet, while some heavy smokers develop lung cancer at a young age, other heavy smokers never develop it, even at an advanced age, suggesting a remarkable variability in the individual susceptibility to the carcinogenic effects of tobacco. We characterized the germline profile of subjects presenting these extreme phenotypes with Whole Exome Sequencing (WES) and Machine Learning (ML).

METHODS

We sequenced germline DNA from heavy smokers who either developed lung adenocarcinoma at an early age (extreme cases) or who did not develop lung cancer at an advanced age (extreme controls), selected from databases including over 6600 subjects. We selected individual coding genetic variants and variant-rich genes showing a significantly different distribution between extreme cases and controls. We validated the results from our discovery cohort, in which we analysed by WES extreme cases and controls presenting similar phenotypes. We developed ML models using both cohorts.

FINDINGS

Mean age for extreme cases and controls was 50.7 and 79.1 years respectively, and mean tobacco consumption was 34.6 and 62.3 pack-years. We validated 16 individual variants and 33 variant-rich genes. The gene harbouring the most validated variants was HLA-A in extreme controls (4 variants in the discovery cohort, p = 3.46E-07; and 4 in the validation cohort, p = 1.67E-06). We trained ML models using as input the 16 individual variants in the discovery cohort and tested them on the validation cohort, obtaining an accuracy of 76.5% and an AUC-ROC of 83.6%. Functions of validated genes included candidate oncogenes, tumour-suppressors, DNA repair, HLA-mediated antigen presentation and regulation of proliferation, apoptosis, inflammation and immune response.

INTERPRETATION

Individuals presenting extreme phenotypes of high and low risk of developing tobacco-associated lung adenocarcinoma show different germline profiles. Our strategy may allow the identification of high-risk subjects and the development of new therapeutic approaches.

FUNDING

See a detailed list of funding bodies in the Acknowledgements section at the end of the manuscript.

Rare variant association on unrelated individuals in case-control studies using aggregation tests: existing methods and current limitations

Over the past years, progress made in next-generation sequencing technologies and bioinformatics have sparked a surge in association studies. Especially, genome-wide association studies (GWASs) have demonstrated their effectiveness in identifying disease associations with common genetic variants. Yet, rare variants can contribute to additional disease risk or trait heterogeneity. Because GWASs are underpowered for detecting association with such variants, numerous statistical methods have been recently proposed. Aggregation tests collapse multiple rare variants within a genetic region (e.g. gene, gene set, genomic loci) to test for association. An increasing number of studies using such methods successfully identified trait-associated rare variants and led to a better understanding of the underlying disease mechanism. In this review, we compare existing aggregation tests, their statistical features and scope of application, splitting them into the five classical classes: burden, adaptive burden, variance-component, omnibus and other. Finally, we describe some limitations of current aggregation tests, highlighting potential direction for further investigations.

Excalibur: A new ensemble method based on an optimal combination of aggregation tests for rare-variant association testing for sequencing data

The development of high-throughput next-generation sequencing technologies and large-scale genetic association studies produced numerous advances in the biostatistics field. Various aggregation tests, i.e. statistical methods that analyze associations of a trait with multiple markers within a genomic region, have produced a variety of novel discoveries. Notwithstanding their usefulness, there is no single test that fits all needs, each suffering from specific drawbacks. Selecting the right aggregation test, while considering an unknown underlying genetic model of the disease, remains an important challenge. Here we propose a new ensemble method, called Excalibur, based on an optimal combination of 36 aggregation tests created after an in-depth study of the limitations of each test and their impact on the quality of result. Our findings demonstrate the ability of our method to control type I error and illustrate that it offers the best average power across all scenarios. The proposed method allows for novel advances in Whole Exome/Genome sequencing association studies, able to handle a wide range of association models, providing researchers with an optimal aggregation analysis for the genetic regions of interest.

Whole exome sequencing characterization of individuals presenting extreme phenotypes of high and low risk of developing tobacco-induced lung adenocarcinoma

Background

Tobacco is the main risk factor for developing lung cancer. Yet, some heavy smokers do not develop lung cancer at advanced ages while others develop it at young ages. Here, we assess for the first time the genetic background of these clinically relevant extreme phenotypes using whole exome sequencing (WES).

Methods

We performed WES of germline DNA from heavy smokers who either developed lung adenocarcinoma at an early age (extreme cases, n=50) or did not present lung adenocarcinoma or other tumors at an advanced age (extreme controls, n=50). We selected non-synonymous variants located in exonic regions and consensus splice sites of the genes that showed significantly different allelic frequencies between both cohorts. We validated our results in all the additional extreme cases (i.e., heavy smokers who developed lung adenocarcinoma at an early age) available from The Cancer Genome Atlas (TCGA).

Results

The mean age for the extreme cases and controls was respectively 49.7 and 77.5 years. Mean tobacco consumption was 43.6 and 56.8 pack-years. We identified 619 significantly different variants between both cohorts, and we validated 108 of these in extreme cases selected from TCGA. Nine validated variants, located in relevant cancer related genes, such as PARP4, HLA-A or NQO1, among others, achieved statistical significance in the False Discovery Rate test. The most significant validated variant (P=4.48×10⁻⁵) was located in the tumor-suppressor gene ALPK2.

Conclusions

We describe genetic variants associated with extreme phenotypes of high and low risk for the development of tobacco-induced lung adenocarcinoma. Our results and our strategy may help to identify high-risk subjects and to develop new therapeutic approaches.

Frequency of Pathogenic Germline Variants in Cancer-Susceptibility Genes in Patients With Osteosarcoma

Importance

Osteosarcoma, the most common malignant bone tumor in children and adolescents, occurs in a high number of cancer predisposition syndromes that are defined by highly penetrant germline mutations. The germline genetic susceptibility to osteosarcoma outside of familial cancer syndromes remains unclear.

Objective

To investigate the germline genetic architecture of 1244 patients with osteosarcoma.

Design, Setting, and Participants

Whole-exome sequencing (n = 1104) or targeted sequencing (n = 140) of the DNA of 1244 patients with osteosarcoma from 10 participating international centers or studies was conducted from April 21, 2014, to September 1, 2017. The results were compared with the DNA of 1062 individuals without cancer assembled internally from 4 participating studies who underwent comparable whole-exome sequencing and 27 173 individuals of non-Finnish European ancestry who were identified through the Exome Aggregation Consortium (ExAC) database. In the analysis, 238 high-interest cancer-susceptibility genes were assessed followed by testing of the mutational burden across 736 additional candidate genes. Principal component analyses were used to identify 732 European patients with osteosarcoma and 994 European individuals without cancer, with outliers removed for patient-control group comparisons. Patients were subsequently compared with individuals in the ExAC group. All data were analyzed from June 1, 2017, to July 1, 2019.

Main Outcomes and Measures

The frequency of rare pathogenic or likely pathogenic genetic variants.

Results

Among 1244 patients with osteosarcoma (mean [SD] age at diagnosis, 16 [8.9] years [range, 2-80 years]; 684 patients [55.0%] were male), an analysis restricted to individuals with European ancestry indicated a significantly higher pathogenic or likely pathogenic variant burden in 238 high-interest cancer-susceptibility genes among patients with osteosarcoma compared with the control group (732 vs 994, respectively; P = 1.3 × 10-18). A pathogenic or likely pathogenic cancer-susceptibility gene variant was identified in 281 of 1004 patients with osteosarcoma (28.0%), of which nearly three-quarters had a variant that mapped to an autosomal-dominant gene or a known osteosarcoma-associated cancer predisposition syndrome gene. The frequency of a pathogenic or likely pathogenic cancer-susceptibility gene variant was 128 of 1062 individuals (12.1%) in the control group and 2527 of 27 173 individuals (9.3%) in the ExAC group. A higher than expected frequency of pathogenic or likely pathogenic variants was observed in genes not previously linked to osteosarcoma (eg, CDKN2A, MEN1, VHL, POT1, APC, MSH2, and ATRX) and in the Li-Fraumeni syndrome-associated gene, TP53.

Conclusions and Relevance

In this study, approximately one-fourth of patients with osteosarcoma unselected for family history had a highly penetrant germline mutation requiring additional follow-up analysis and possible genetic counseling with cascade testing.

Identifying individual risk rare variants using protein structure guided local tests (POINT)

Rare variants are of increasing interest to genetic association studies because of their etiological contributions to human complex diseases. Due to the rarity of the mutant events, rare variants are routinely analyzed on an aggregate level. While aggregation analyses improve the detection of global-level signal, they are not able to pinpoint causal variants within a variant set. To perform inference on a localized level, additional information, e.g., biological annotation, is often needed to boost the information content of a rare variant. Following the observation that important variants are likely to cluster together on functional domains, we propose a protein structure guided local test (POINT) to provide variant-specific association information using structure-guided aggregation of signal. Constructed under a kernel machine framework, POINT performs local association testing by borrowing information from neighboring variants in the 3-dimensional protein space in a data-adaptive fashion. Besides merely providing a list of promising variants, POINT assigns each variant a p-value to permit variant ranking and prioritization. We assess the selection performance of POINT using simulations and illustrate how it can be used to prioritize individual rare variants in PCSK9, ANGPTL4 and CETP in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) clinical trial data.