Skip to main content
Fig. 4 | BMC Genomics

Fig. 4

From: Assessing characteristics of RNA amplification methods for single cell RNA sequencing

Fig. 4

Single-cell RNA sequencing accuracy. a Average pairwise correlation of diluted replicates with bulk HBR or UHR. Zeros treated as missing values. Error bars indicate ± 2 s.e.m. bc Relationship between % unique alignment and similarity with bulk HBR or UHR. “r” indicates Pearson correlation of x and y. (B) 10 pg. (C) 100 pg. df Distribution of fold deviation across genes. Wide boxes represent measurements in individual replicates. Narrow boxes represent average measurements across replicates. Y-axis was truncated for visualization and 99%ile values for wide boxes are in panel descriptions below, ordered to match plot. “*” indicates significant difference between individual and average measurements (Wilcoxon rank sum test of greater fold deviation in average measurements, p < 0.05). (D) aRNA. 99%ile values: 2066; 1442; 514; 718. (E) SmartSeq Plus. 99%ile values: 877; 770; 553; 598. (F) NuGen. 99%ile values: 2332; 1766; 784; 937. gj Enrichment of biophysical traits among underestimated (low) and overestimated (high) genes with respect to remaining genes. See Methods . Plot notation and statistics are as in Fig. 3c–f. (G) Median difference in gene length. (H) Relative risk of containing an internal A-hexamer. (I) As G, for % GC content. (F) As G, for strength of local secondary structure. km Density scatter plots of normalized read counts in individual 10 pg. replicates vs. expected number of input molecules. See Methods . Red indicates high density. Solid line indicates expected read count and hashed lines indicate ± 2 fold. (K) aRNA. (L) SmartSeq Plus. (M) NuGen. np Density scatter plots of average normalized read counts vs. number of input molecules. (N) aRNA. (O) SmartSeq Plus. (P) NuGen. Gene filtering: D–F and K–P considered computationally unambiguous genes and excluded gene detection outliers. D–J considered genes with greater than 95% probability of presence in a diluted replicate. Abbreviations: confidence interval (CI); kilobases (kb); kilocalorie (kcal); mole (mol)

Back to article page