Fig. 1

Overview of Droplet-CirSeq. a Droplet-CirSeq workflow. Genomic DNA was sheared into fragments shorter than half the length of the sequencing read and then denatured into single-stranded DNA molecules that were circularized using single-strand DNA ligase. After eliminating the linear DNA using DNA exonucleases, the circularized single-stranded DNA was used for RCA (rolling circle replication). The circularized DNA and RCA reaction mix was added to a RainDrop Source chip to produce water-in-oil emulsion droplets. Generally, approximately 5 ~ 10 million droplets formed in approximately an hour in a 50 μl volume. The droplets containing RCA mix were allowed to continue to react for 4 ~ 16 h in order to amplify enough DNA for standard NGS library preparation in the following steps. Please note that the insert size of the standard NGS libraries must be larger than twice the length of the original circularized DNA to avoid sequencing the same DNA copy twice instead of sequencing two independent-amplified copies. b Error correction. Here is an example to explain the error correction strategy. Multiple copies of the original circularized DNA were examined in every PE read. “A” (red color) represents the base, which may have errors generated during PCR or sequencing. There will be three cases present in the sequencing result: AA (case 1), no errors; AB (case 2), one read error; and BB (case 3), two read errors. B stands for T/C /G. In the following bioinformatics analysis, Case 2 and Case 3 will be filtered except when BB has the same bases, such as TT, GG, or CC (false positive)