Fig. 2

Sequencing saturation for gene quantification. a Percent of all possible UMIs observed in a sample (sequencing saturation) as a function of the total number of aligned reads (library size). The black dotted line shows the theoretical sequencing saturation assuming no PCR bias or sequencing errors in UMIs (binomial distribution). The colored dotted lines show sequencing saturation modelled by a negative binomial distribution fitted to the observed values, separately for each sample type. One hundred percent means 65,536 unique UMIs. b Association of the average number of reads per UMI and total number of aligned reads (library size). The black dotted line shows the theoretical relationship based on the binomial distribution. Colored solid lines show linear regression fitted to observed values for each sample type