describe an advance in imputation quality afforded by the use of a newly designed genotyping array used in conjunction with the largest reference panel available so far. A slightly modified version of this array is being used to genotype the remaining samples in UK Biobank. As evidence of the usefulness of these data, we describe novel insights into the genetic architecture of airflow obstruction and smoking. Specifically, we show that there are shared genetic causes of airflow obstruction between smokers and non-smokers, consistent with the limited evidence for gene–smoking interactions described so far. We show that the genetic determinants of low FEV1 in individuals without asthma are also informative in individuals with asthma. We report new loci associated with extremes of FEV1 and COPD, including evidence that a genomic region of complex structural variation has an effect on lung function and airflow obstruction in the general population. Our novel signals implicate epigenetic mechanisms as contributors to lung health. These findings, taken together with previous findings, will help define pathways underlying predisposition to development of COPD and smoking behaviours. A full understanding of the biological mechanisms underlying these genetic associations will improve our understanding of the pathophysiology of COPD and smoking
