Evaluation of the performance of hashtags in demultiplexing: (A) Evaluation of global demultiplexing efficacy (GDE) in each pool. Bar diagram showing percentage of hashtag singlets within all cells in a well. The first five bars are the calculations of GDEs of wells A–E, combining two hashtags within a sample as one. The approximated mean GDE is noted on the boxes covering the lower parts of the bars. (B) Evaluation of individual hashtag efficacies (IHE) for each hashtag used in each pool shown as dot plots accompanied by overlaid box plots. Pools and the number of hashtags (in brackets) used in them are marked on the x-axis. The hash (#) signs are abbreviations for “hashtag.” For A and B, the expressions of GDE and IHE, respectively, are noted in the gray boxes at the top. (C) Cartoon Venn diagrams to illustrate the relations between overlaps and the availability of exclusive parts of intersecting sets. (D) A closed-form expression to calculate exclusive parts of sets (E), given that we know x, y, and n (see text) (Supplemental Text S1). (E) Charts showing calculations of E, for the given value of y and x (shown at the top of the charts). The different values of x are at the top row, preceded by a ∩ (intersection) sign, and n is the first column in the charts. (F,G) Estimation of x and y in single-cell data. Three charts are for the pools with four hashtags; the first columns indicate the specific hashtag for which the overlap is calculated, and the second columns are the estimated values of x (F). Two bar plots at the bottom are for pool F, namely, 12 hashtags (G). The title above the charts/bar plots, such as “F, #4,” indicates that x and y are estimated for hashtag 4 in well F. Below that, the percentage of exclusive signals, namely, singlets for that hashtag, is noted. The bars or the second columns of the charts are the values of x, as estimated directly from data (see Methods). Numeric estimates of y are noted immediately below the bars or the chart entries (i.e., the corresponding x values).