(a) Bioanalyzer traces of libraries shown in Fig. 1c, demonstrating their different length distributions. (b) Sequencing the libraries shown in (a) on an Illumina HiSeq3000 results in higher fractions of 5’ UMI reads than when the same libraries are sequenced on the Illumina NextSeq500 (shown in Fig. 1c) (n = 16 HEK293FT cells per condition). Sequence machine biases are likely fragment length related. (c) Enrichment of 5’ UMI containing reads after tagmentation with a linear PCR step (Forward pre-amplification PCR primer) of either 2 of 4 cycles, before adding index primers and index PCR (n = 16 HEK293FT cells per condition). (d) Increased UMI containing reads with addition of custom i5 Illumina Index oligos targeting the 5’ tag during index PCR (n = 96 Fibroblasts per condition). (e) HEKF293FT cell bioanalyzer traces showing the effect and ability of the template switching oligo priming in PCR in absence and presence of varying amount of forward PCR primer. (f) UMIs detected at 100.000 UMI-reads at varying forward PCR primer concentrations with and without the presence of template switching oligo in PCR reaction (n = 48 HEK293FT cells per condition). (g) Number of genes detected from 100.000 UMI containing reads with increasing amount of forward PCR primer with or without the presence of the template switching oligo in PCR reaction (n = 48 HEK293FT cells per condition). Significance in (f and g) was evaluated by two-sided t-tests, indicated on the figures. Boxplots denote median and first and third quartiles. Whiskers indicate the most extreme data point within 1.5 lengths of the box.