Genomic predictors of smoking and lung cancer risk

点击翻译按钮获取中文摘要

Abstract – Project 1 Lung cancer (LC) is the leading cause of cancer. Lung cancer is among the best examples of a disease resulting from a complex interaction between environmental exposures and genetic factors. Germline genetic findings make an important contribution to the definition of high-risk individuals and provide key insights into LC etiology. LC genome wide association studies (GWAS) have identified informative loci that have influenced our approach to tobacco control and provided new insights into tumorigenesis. We have identified 24 loci with involved in cancer susceptibility. However, the interplay between inherited susceptibility and effects from demographic and environmental factors has not been elucidated. We hypothesize that genetic variation influences both smoking behaviors and cellular processes that jointly lead to lung cancer development. Our specific aims are: Aim 1 of this project will characterize the contribution of common genetic variation to lung cancer etiology. We will analyze GWAS of lung cancer from 47,506 lung cancer cases and 63,687 to identify factors influencing lung cancer risk according to histology and host-characteristics. Mechanisms by which these variants influence cancer risk will be explored using eQTL based procedures and through annotation of existing databases. Aim 2 will investigate uncommon genetic variants for LC susceptibility. We will analyze exome germline sequencing information from over 2,500 lung cancer cases and 2,000 controls use these for imputation and rare variant analysis. Aim 3 will identify genetic effects on smoking behavior. For this aim, we will integrate the large-scale genetic studies we have conducted with extensive phenotyping performed through the lung cancer cohort consortium (LC3) and project 2 to identify the specific impacts of SNPs on smoking behaviors. Aim 4 will characterize joint effects of environmental and genetic interactions on lung cancer risk. This extensive data and information from cohort studies we have assembled will allow us to model joint effects of smoking and genetic factors on lung cancer risk over time. Supported by the biostatistical core we will perform mediation analyses to partition risk among multiple smoking phenotypes and genetic factors. We will also use the genetic data to perform Mendelian randomization to evaluate the relevance of biomarkers in predicting lung cancer risk, to assist project 2.