Fig. 2

Sequence performance of Quartz-Seq2 with molecular biological improvements. a Improvement of poly(A) tagging efficiency. The relative DNA yield in various poly(A) tagging conditions using purified first-strand cDNA from 1 ng of total RNA. T55 buffer as the terminal deoxynucleotidyl transferase (TdT) buffer and the temperature condition “Increment” for the poly(A) tagging step improved the cDNA yield of whole-transcript amplification. Buffer compositions are indicated in Additional file 6: Table S5. QuartzB represents use of a Quartz-Seq-like buffer as a positive control, in accordance with the approach described in the original Quartz-Seq paper. Finally, we quantified cDNA yield (300–9000 bp) and byproduct DNA yield (50–300 bp) using a Bioanalyzer (Agilent). The presented p value was obtained using two-tailed Welch’s t-test. b Reverse transcription efficiency with serially diluted RT enzymes. The x-axis represents the average relative RT qPCR score from ten genes. Detailed concentrations of RT enzymes are presented in Additional file 1: Figure S7. c, f–h Comparison between Quartz-Seq2 in the RT25 condition and Quartz-Seq-like conditions regarding sequence performance. c We analyzed 384 wells with 10 pg of total RNA and used approximately 0.19 M fastq reads on average per well. We show the UMI count and gene count in box plots. d A scatter plot between the mean of gene expression and the variability of gene expression with 10 pg of total RNA in 384 wells. Red lines represent the theoretical variability of gene expression in the form of a Poisson distribution. e Gene expression reproducibility between bulk poly(A)-RNA-seq (1 μg of total RNA) and Quartz-Seq2 (10 pg of total RNA, averaged over 384 wells). f Dispersion of gene expression. The x-axis represents gene expression variability. g Reproducibility of gene expression for internal gene and external control RNA. h Accuracy of gene expression for internal gene and external control RNA