To investigate the effect of rare coding variants on smoking behaviour, we studied 346,813 participants (of which 324,851 were of European ancestry) from 61 cohorts (Supp. Tables 1 and 2) at up to 235,116 SNVs from the exome array. As we had access to UK Biobank, we also interrogated SNVs present on the UK Biobank and UK BiLEVE Axiom arrays to identify additional associations across the genome beyond the exome array. To our knowledge, these datasets are an order of magnitude larger than the previous studies [6], and constitute the most powerful exome-array study of smoking behaviour to date.
