Skip to main content
. 2017 Sep 14;8:1157. doi: 10.3389/fimmu.2017.01157

Figure 2.

Figure 2

Forming unique molecular identifier (UMI) groups by sequence clustering increases the number of sequences passing filter. (A) Allowing one error in the UMI sequence increases the sequences deriving from three or more raw reads compared to the grouping of UMI by read identity. In addition, this leads to a decrease in the number of ambiguous calls (N nt) during consensus assignment. (B) Sequences deriving from UMI groups by sequence clustering have higher UMI counts.