Approved research

Translating genetics of nicotine dependence to pulmonary function

Lay summary

We aim to identify genetic associations with nicotine dependence by conducting GWAS meta-analyses and expand our findings to smoking-related health outcomes, e.g., pulmonary function. We have assembled 16+ GWAS study samples with Fagerstrom Test for Nicotine Dependence data, totaling N>45,000 of European and African ancestry. We propose using results from GWAS of UK Biobank data to replicate genetic associations with heavy vs. never smoking and translate our associations to pulmonary function. We also propose using the data to perform GWAS analyses that condition on known SNPs and incorporate SNP-by-SNP and SNP-by-gender interactions. Nicotine dependence is one of the strongest predictors of failing to quit cigarette smoking. Results of this genome-wide study may be used to better understand the genetic risk factors for nicotine dependence, smoking cessation, and other smoking behaviors and outcomes and ultimately reduce the burden of smoking's health consequences including lung cancer and chronic obstructive pulmonary disease. Our study aims to identify and replicate new genetic variants associated with nicotine dependence. Using the largest collection of nicotine dependence samples with genome-wide data assembled to date, we are identifying new genetic variants by using the standard genome-wide association study approach and further by testing interactions with gender and with nicotine receptor and metabolism genes that are already known to influence a person's risk of becoming nicotine dependent. We propose using the UK Biobank data to take the same analytic approaches in assessing genetic variant associations with heavy vs. never smoking and pulmonary function (e.g., FEV1), an important smoking-related trait. We propose using the participants with GWAS, smoking, and pulmonary function measures available (e.g., FEV1). GWAS of smoking and pulmonary function on these participants (N~50,000) were published in 2015 by other researchers using UK Biobank data (Wain et al. Lancet Respir Med, PubMed ID 26423011).

Scope extension:

We aim to identify genetic associations with nicotine dependence by conducting GWAS meta-analyses and expand our findings to smoking-related health outcomes, e.g., pulmonary function. We have assembled 16+ GWAS study samples with Fagerstrom Test for Nicotine Dependence data, totaling N>45,000 of European and African ancestry. We propose using results from GWAS of UK Biobank data to replicate genetic associations with heavy vs. never smoking and translate our associations to pulmonary function. We also propose using the data to perform GWAS analyses that condition on known SNPs and incorporate SNP-by-SNP and SNP-by-gender interactions. When modeling pulmonary function and its decline over time, we propose a comprehensive set of GWAS models, including a standard model (i.e., no interactions) and interaction models that embed SNP interactions with smoking, gender, other SNPs, and/or nutrition.

We propose to expand our scope using the recently-released metabolomics data to augment the knowledge gained by the GWAS approach. The release of 249 metabolites measured in 120,000 participants represents the largest single metabolomics dataset to date. We intend to combine our analyses of genome-wide association data with the metabolomics data to further understand the genetic underpinnings of pulmonary function and smoking-related traits. Metabolomics has already been shown to be associated with smoking-related traits in much smaller studies (PIMD34650005, 30939782), and we will use this larger study to identify new metabolite associations and independently replicate previous ones. We intend to combine the metabolomics data with the genetic results obtained in our initial study to identify those associations that are likely to be mediated by metabolic perturbations. Metabolites that replicate previous associations or identify new associations with smoking-related traits will also be targets for additional genetic interrogation, as metabolites representing intermediate phenotypes may provide additional genetic variant associations due to the larger expected association size. Combining available UK Biobank data in these domains with other publicly available large-scale data using multi-omics pathway integration may yield yet another level of biologically relevant information about the causes and consequences of smoking-related phenotypes.