UKB Exome Sequencing Processing

The 469,809 UKB exome sequences were processed at AstraZeneca from their unaligned FASTQ state. A custom-built Amazon Web Services (AWS) cloud computing platform running Illumina DRAGEN Bio-IT Platform Germline Pipeline v.3.0.7 was used to align the reads to the GRCh38 genome reference and to perform single-nucleotide variant (SNV) and insertion and deletion (indel) calling. SNVs and indels were annotated using SnpEFF v.4.3 (ref. ^{62 (link)}) against Ensembl Build 38.92 (ref. ^{63 (link)}). We further annotated all variants with their gnomAD MAFs (gnomAD v.2.1.1 mapped to GRCh38)^{64 (link)}. We also annotated missense variants with MTR and REVEL scores^{16 (link),29 (link)}. The AstraZeneca pipeline output files including the variant call format files are available through UKB Showcase (https://biobank.ndph.ox.ac.uk/showcase/label.cgi?id=172).