Fig. 1

Effect of errors on the Duplex Sequencing procedure. Here input DNA is sheared and barcodes are ligated to the ends of the DNA molecules (colored rectangles in Barcoded DNA). After paired-end library preparation and sequencing each original molecules gives rise to multiple reads (Paired-end reads pane). This process also inadvertently generates sequencing errors represented by yellow rectangles and red circles. The yellow rectangles and red circles are used to depict errors arising inside read compartments corresponding to original DNA and adapters, respectively. Reads are then grouped by barcode to produce “families”. In this example each family is required to contain at least three reads. As shown here one of the reads contains an error (red circle) within the barcode. The error makes this particular barcode different from others. As a result it cannot be added into the family and remains a singleton (the error correction algorithm described here was developed specifically to correct such errors and allow singletons to be joined with their respective families). Each family is subsequently reduced into a Single Strand Consensus Sequence (SSCS) and each respective SSCS is merged with its counterpart from the opposite strand to generate Duplex Consensus Sequences (DCS)